I have spent some time testing the FoCal software. Because I made some comments earlier in I would add a summary of my experience/opinion.
There are a couple of issues. It is the responsibility of the tester to insure the camera sensor and target are as parallel as possible. While I've found it isn't ultra critical (an inch or two), a few inches off line does affect the outcome. This isn't too challenging to over come, placing a small mirror on the target when positioning the camera, allows one to fine tune the camera position quite accurately. Not as elegant as the LensAlign, yet not too difficult to overcome. Printing your own target means you need a good printer with MK ink. I used my 11880 on Epson premium presentation paper at 2880. The resulting target was sharp and worked well.
At this point I've worked extensively with a d800e, a 5D Mark 3, and a 5D Mark 2. I have a d700 and a D4 coming in, but I don't think they will change my perspective.
As far as the d800e, I don't think the software can handle the moire when analyzing the target. On nearly every lens the setting chosen was off by as many as 7 or 8 points. It sees the moire as lack of sharpness. However, by clicking on the focus points in the analysis, it was pretty easy to find which was the sharpest, so I used the software in the manual mode, and I think it was easier to use than when I tried the lens align. The issue with d800e for me so far is I just haven't tested any lenses that can resolve anywhere close to the sensor wide open, soeverything is blurry. With the lens align it's difficult to decide which area is "less" blurry. I found it easier to calibrate the d800e by shooting a target which had some other information such as images which didn't show the moire problem. Also it was interesting to see how quickly diffraction affected moire (by f/11 with 3 different lenses most of the moire was gone).
The 5D Mark3 was much easier to use, but it it isn't totally automatic. Some of the canon's are fully automatic, start the test, come back 5 minutes later. This was the case with the 5D mark 2. What I did find was using the recommended setting wasn't always the setting I would choose. When you examine the data you can see the results of any shot by clicking on the diamond and visually compare it to the selected setting/shot. With the 5D Mark2 and the Canon 100mm macro, the software recommended a setting of -3, but visually it was obvious that a setting of -6 yielded the most accurate focus. The difference was at -3 the point spread was much tighter, where as at =6 there were a couple of results worse than -3 but several that were better. So one caveat I would make is that after you complete the test, you visually examine the charts and compare the various points with the recommended/selected one. On the 16-35 mm, the software nailed it (and it was significant, -13). It also did so on the 70-200. The 100mm Macro is a pretty old lens, and showed by far the most deviation in focus.
I found the f/stop range test very interesting, and it was useful in determine the minimum/maximum values for optimum sharpness. On the 5D Mark 2, all 3 lenses were better stopped down, and at f/11 diffraction had very little impact. After that it became obvious and by f/22 things were a complete blur. I'll be testing this out with the d800 as soon as I can get my hands on one.
As far as the licensing, personally I find it reasonable and not an issue. You are allowed to enter 5 serial numbers for cameras, and you are allowed to make changes to your serial number list 10 times. The change is a matter of logging onto their website, making the changes to the numbers, and then copying and pasting the resulting code back into the software. It only takes a couple of minutes, and seems a very reasonable method to prevent abuse (say a camera club buying one copy and then calibrating everyone's camera.) For around $400, you can license it to a CPU rather than camera bodies, and then there is no limit to the cameras. I believe just like bodies, you are allowed to change the link to the CPU a few times.
I ran the latest beta version of their software on a Mac using Parallels/Windows 7. I had no issue with the USB cable supplied with the d800e or with USB for the Canons.
The concept seems well thought out, the feedback is excellent and the algorithms to analyze the data seem to work pretty well for the most part. When they don't the visual feedback is useful and works. I found I could calibrate the 5D Mark 2 in less time than the Lens Align, and personally believe I have more confidence in the results. Most where quite similar to the same setting as with LensAlign, but most were a point or two different.
All of this is of course just my humble opinion, your mileage and opinion certainly might be different.