[gumpzilla]

I had missed that this pays out two, so mea culpa on my earlier comment that suggested otherwise.

Assigning equity results 2/1/0 - not entirely right, but pretty close - let's say pushing wins the blinds 100% of the time and nothing else. Our equity following pushing is 1.33. Getting into an all-in confrontation gives us 1.68 when we win, 0 when we lose. So, if we suppose that limping means we get pushed over by any two, we need to be confident that we'll win 1.33/1.68 = 79.2% for limping to be superior. Equity of QQ vs. a random hand: 79.9%. Consider that pushing will do slightly less well than just winning the blinds if the player is calling only with hands that beat us, as you suggested, and it seems to me like the numbers suggest that limping might be very slightly superior. It's close, at least.

EDIT: Going back and putting in .65/.35 to be a little more rigorous about it, it pushes the desired victory percentage up closer to 81%. Wacky.