Data has been taken by Fabrice (exposure sequence from 670 to 683). Neon data is of good quality. Unfortunately, Argon data was taken with same exposure times (exptime=200) in both focus and defocus - this means that data is overexposed and often saturated in focus, while having very low signal-to-noise ratio in the defocus. Argon data would have been particularly useful, as it spans the detector better than neon data.
I have run the first batch of the analysis on different spots/donuts. I have done this n the image which have been ``corrected'', i.e., for which detector has been moved from the nominal focus position in a position which Fabrice has calculated to be the best given the issues as SM2. The analysis has been successful for 20 spots/donuts. The unsuccessful donuts were either oversaturated or very near the edge. There are few which I will rerun, which failed due to trivial mistake in my code. I have placed a figure with singular example of this very quick analysis in body of the ticket. 
I am proceeding with the quick analysis of the rest of the neon data. It will be useful do the same on Argon data, if we are able to get the data.

I have sent the rest of the neon data, and the argon data (without data in focus) to be analyzed on the cluster.

Given that we have received only limited data, the full analysis can not be done without further considerable effort. If LAM insists I can spend more time on this, but I believe that the cost in time and effort is larger than the gain. It would be much easier with proper data, which might be done after LAM is back in September. Some useful insight has been gained, especially in seeing how the defocus pupil looks on SM2.