This is indeed a truly great fix and I can't commend you guys enough on its design. Well done!
I've used it a handful of times and it seems to be working on the 2 units I retained.
I must say that I will now ALWAYS perform the heatgun/reflow first, followed by the method described here. Last night I had a stock RRoD 0001, which I reflowed first then tested out very quickly without the heatsinks attached. Just verified I got the green lights and switched it straight off again. Then went ahead with the rest of this clamping method word for word.
This means that the reflow is the actual fix, and the clamping becomes the preventative measure. Much better than the clamping being the fix in my opinion.
One unit was handed to me a while back as a "bag o bits" after the previous owner's failed attempt at fixing it. I was literally handed a carrier bag. Today, with this method it's being going strong for 1 month which has included several hours each day in my quality assurance department (OK, it's the kids playing lego stars wars - they don't want to stop.)
One other idea I'm toying with (but haven't needed to try out yet) is if the initial reflow itself doesn't fix it, then I'd guess that one potential reason for this is that a solder joint is broken badly enough (or the board is warped badly enough) that the 2 parts of the broken joint are too far apart to touch. In which case I'd considered reattaching the heatsinks very lightly with 10mm screws (no thermal paste) and re-doing the reflow. Maybe also include the washers on the DRAM too, to try and get the board as flat as possible during the reflow.
Anyway, that's my experience.
Thanks again chaps - great work.