Hi,
As far as I know it isn't a standard method of collimation, I had to develop it after my Cheshire showed the scope as collimated but my subs showed severe coma and distortion across the upper right hand third of the fov.
This drawing sort of explains the idea, sorry about the quality!
The main concept is that coma and *secondary* light-cone misalignment are related when primary misalignment is removed. Image coma and vignette are in the same places as each other, when the primary is collimated relative to the secondary. In other words, the method can only adjust the secondary and therefore requires use of a Cheshire to keep the primary constantly aligned with the secondary throughout the process.
To get a flat, I simply take a quick shot through the scope with my laptop screen open in Notepad, and then take the result into GIMP or StarTools to do a full desaturation and a savage clip/stretch. This was the result from the scope before any secondary adjustments, bear in mind that at this point it was "perfectly" collimated according to the Cheshire *and was rotated/centred almost perfectly, too*.
The dark gradients were indeed in the same place as the bad coma, and they fell in the position I had predicted.
As shown on the diagram, I then worked out which direction on the photo was which direction for the secondary, then started to adjust. Throughout the process the corrections were perfectly logical and followed my 'hypothesis'. Note that the same stretch has not been applied to each flat, I altered the black point etc to optimise each one for the section I was trying to clean up. Shots 1-3 below show me getting the vertical (horizontal for the secondary) axis correct, then numbers 4-5 show me rotating the secondary into place.
After a bit of work I got this ^.
Most of the adjustment I had to make was rotation on the secondary, (up is to the right on the photos, so it was drooping a bit) and a little bit of centring with the vanes; I made sure that the primary was correctly adjusted after each movement of the secondary before I took the flat to estimate mis-collimation. I was then able to get almost entirely coma-free images.
Hope this helps
John