For what it's worth I have found a way around this.
Basically you just want to break the weld between the surfaces before lifting the HSF off the processor. I find a slight twisting motion can break the weld if present.
I have heard of AMD processors being pulled right out of the socket by removing the HSF way before it happened to me. So IMO its one of those things that doesn't happen until it does,....unless you know how to avoid it,...which I now do.
With ZIFF the pins are tenuously held once the leaver is closed. With LGA there is a metal bracket that clamps down on the package heat spreader and is locked into place.
The only way I see an LGA bracket failing like this is if the socket / assembly wasn't attached to the motherboard properly, the locking mechanism failed or was never engaged properly.
On this board (A Dell system) I don't think I could twist the heat sink is nice enough for me to use on some LED projects nice and thick but it was impossible to twist but that would be the normal way I would of did if I could not break the seal.
The dell uses this large plastic base in which the HSF sits in not allowing it to twist.
