Okay, my initial input. When buying tech, we are usually buying it to solve a specific problem. Generally speaking the features that help solve problems are -- and please free to add to the list.
First let me add some background data: The ends of the camera with movements are referred to as "standards." In view camera parlance it was for the frames between the bellows. The rear standard holds the sensor, the front standard holds the lens, the bellows connects them and at the same time allowing movement on the standards while keeping the space between them light-tight. Camera movements can be on either standard or both in varying degrees. Tech cameras generally replace the bellows with a rigid frame for more accurate parallelism between the front and rear standard than a view camera can manage. Because of this, it is a more difficult manufacturing process to get full movements at both ends, so other compromises need to be made. This is why certain "hybrid" cameras are available, being half tech body and half bellows camera; the goal being to gain and maintain advantages of both platforms. Hence the issue regarding which features -- the more there are, then generally speaking the higher the cost to build and buy. Note that which standard the movements are on *IS* critical for solving specific issues.
Main features:
1) Superior glass available, usually better on resolution and/or distortion than DSLR glass.
2) Superior precision in alignment between sensor and lens.
3) Shift and rise movements. Very helpful in composing while maintaining visually correct 3D to 2D geometric projection onto the sensor. If these movements are at the lens standard, perspective is altered when making the movements. If it is on the rear or sensor standard, then the movements maintain original perspective. The latter is useful if one plans to shift-stitch for higher resolution.
4) Tilt and Swing movements. The only way to alter the PoF (Plane of Focus) off the perpendicular to the sensor. This allows for extending DoF beyond what is possible from just stopping down the aperture. If on the rear standard, they impart geometric distortion in the 3D to 2D projection, but this can be used to visually adjust aspects to an image to make them more appealing, like decrease the effects of an angling away wall or sides of a box on a product.
5) Other aspects to consider are build-quality, cost, breadth and depth of available accessories, manufacturer and dealer support, and overall product availability. Some of us are going to be more willing to give in on some of these than others, so in the end the end they do become valid points to consider.
~~~
To be clear, I have owned and a used a plethora of view and technical cameras. They all had features I liked and they all were missing features I would have liked. A few had features I liked but were poorly implemented, while others had features I had limited need of yet were superbly implemented. And the reality is I could generate images with all of them that were superior to similar images I could generate with a DSLR. Another reality is the net improvement in my PRINT-level output detail and technical superiority was relatively minor; IOW if I had not known what I could have done or had a side-by-side to compare using a bit of rise - shift - tilt - swing, I would probably not have cared very much. So my honest view is that going with tech adds the last tiny level of technical improvement and not any quantum level of performance or creativity enhancement. So don't expect miracles when buying into it, as it still requires a talented operator to generate superior results. Moreover, tech cams require methodical process to use well and as such are infinitely easier to totally screw up with compared to an MF DSLR, so the down-sides to their use are often more severe and more plentiful than the gain-sides. So, with all that said,
~~~
My ideal camera would have all movements at both ends and be in the rigid design of a bellows-less tech camera. This camera does not exist, so I needed to make some compromises. First I wanted rigidity and repeatable zero for all movements. I also wanted relatively compact and hand-holdable. This let out most any camera with a bellows between the standards as they simply are not as rigid as a solid body, nor as compact. Next I wanted shifts and rise at the rear so I could do perspective correct flat stitching. This let out cameras with only one or the other of those movements on the front standards. Next I wanted tilt and swing up front for altering PoF, and I wanted the movements available for my shortest lenses. This limited my choices significantly, primarily because I had already eliminated bellows cameras for rigidity, and having a bellows is the easiest (and cheapest) way to achieve these front tilt and swing movements.
After the above conditions I was left with two systems to contemplate. The Sinar would have been in the running, except it has a permanently attached sliding back which leaves it impractical as a "compact" or even hand-held choice. (More on sliding backs in a minute.) Alpa was out because it's only tilt option is an adapter that is only good for 80mm and longer lenses, and I wanted to be able to have tilt on my wides. This left the Cambo WDS and the Arca RM3d. The Arca only has tilt or swing, not both at the same time. The Cambo with a lens mounted in the TS mount has both, but only down to a certain 35mm lens. (Note I even wrote a review here showing why both are beneficial:
http://forums.getdpi.com/forum/showthread.php?t=13238) This latter point was not too disturbing, as with hyper-wides like the 23 I am lusting after, tilt becomes less necessary --- but nonetheless it was a half-tic against the Cambo. This left the Arca, but I cannot have tilt and swing at the same time, which was a full tic against it. In reality, I only ever need tilt and swing together very rarely, but when it is needed it's defintiely convenient to have. So I went back and looked at Cambo. One little nit I had with the Cambo when I tested it was the overall build quality and implementation of the TS mount. It works, but the zero indent for the lens swings and tilts is not very positive, and I noticed it had maybe a 1/4 degree of "slop" in the stop. While a quarter degree doesn't sound like much, it can definitely throw off PoF, especially in wider lenses used at longer focusing distances. Since I am primarily a landscape shooter, this added another half-tic off the Cambo. By contrast, the Arca's tilt zero detent is very positive and precise, and the adjustment has more granularity as well, allowing for very precise settings. In the end, the Arca's added level of precision won out over the Cambo's T&S feature, but the reality is I could easily make either system work for me. Note that the above is about ME and for MY USES. Your needs and uses may vary, so I advise everybody to investigate all of the available systems and make your own decisions carefully and based on the features available that help achieve your goals and desires.
~~~
A note on cost. All of these systems are priced at a point where a healthy investment is going to be required. So my advice is to pony up to buy the system you really want the first time around. If you find a bargain price on an outfit, it is usually a less popular brand, or has a field of older glass packaged with it as an "not separable" kit -- and a lot of the first and second generation "digital" lenses do not have adequate coverage or performance to work with many of the last three generations of digital back technology. I would also advise build your kit slowly. Buy the body that best suits your needs, and then add the minimal few lenses you have to have to make the kit work as finances allow. Most of us will be supplementing an existing DSLR platform, so other focal options remain. Finally, selling off less popular brands and older glass can be difficult -- and here difficult almost always means you will take a bath financially getting out of it, so be careful how and what you buy. Even if you think you are buying the best, be prepared to sell at half what you paid for it, regardless of how good you think your original buy was; the market for digital MF tech is growing, but is still very small and the laws of supply and demand are definitely at work. Moreover, because of the entry costs, MF digital tech leans toward the wealthier shooters. And wealthier shooters have the ability to buy new, and usually will unless the used is exactly what they want and priced at too good to pass up. This pit can be deep, and you want to enter without slipping!
~~~
A note on sliding backs: Alpa claims they cannot be manufactured to tight enough tolerances to be reliable for their cameras. At first this sounded like marketing speak to me, but then I did some thinking and some math. A sliding mechanism definitely needs tolerance between the sliding pieces in order to move. Over the length a slide has to move for MF, I assumed this be on the order of .002 inches or about 50 microns. It could be less, say .001" or 25u. I can confirm that a shift in focus of 10u VISIBLY moves the focus point on my 40mm HR if it is focused at distance like hundred or so meters. Granted, aperture DoF will accommodate much of this, but the reality is, the lack of precision creates a visible shortfall. Whether it is practically significant or not, I will leave up to you to decide. Personally, I see it a marginal "benefit to loss" tradeoff, meaning it probably helps as often as it hinders. However the reality for me is my back has a good enough 100% review (and a focus mask tool and even a marginal but good enough live view) to confirm in the field my focus point is where I want it, and so a sliding back becomes less necessary, again for ME. Ultimately this is a decision you need to make for your chosen combo of camera and back.