I used the razor blade method and took only a few min to pop it off.
Ok I cleaned it up as best as I could without scratching the pcb. I took a fine Sharpie marker and traced the outline, I used the ArctiClean 2 part cleaner.
I noticed that with the lid off the pcb was convexing upwards, so anyways after all was clean I used Gelid Extreme paste and did a real thin even spread over the die.
Dropped in into the socket, took the IHS and used my marking as a guide and then tried to clamp it in place but the dam thing would slide downwards from the sockets clamp.
A few adjustments and I got it bang on. I then put the water block on and tightened it all the way down as normal.
Now the lid's glue can cure as is, and I am anal about perfection but lacking a proper clamp to let the silicone dry I did what I thought was best.
And it posted.. I saturated the loop for 2 15 min Realbench stress runs for a base line and after the remount I did the same.
Only difference is a +5c ambient room temp with the dam sun poking through the window but all in all I'm happy.
before de-lid
66-67-64-65c
After de-lid
59-61-56-58
So a difference of -7c -6c -8c -7c
Could it have been better? I'm sure it could've if I used liquid metal but I used what I had laying around. Plus a few heat cycles I'm sure the temp's will be lower yet.
BEFORE
AFTER