OK here's an idea.
I can test run the shim using inexpensive thermal paste.
The idea is, i fit the shim over and put some paste on the core. Then i take a perfectly flat surface (like the bottom of an ALPHA or my lapped WB) and i sweep it over the processor. If all the paste is swept i can consider the shim is slim enough.
If there's some residue i can lap the shim a bit more, then try again...
Overall the shim could transfer heat from the ceramic surface to the base of the cooler. This way heat is spread all over the base (which belong to us anyway

) and the result is a greater contact area.
Some HW sites tested perfect shims and had better (=lower) temp readings.
Now if you ppl have tried 'perfect' copper shims and had disastrous results.. well... just tell me why. And no 'copper shims are evil' plz

i know you can have quite valid reasons...