Not sure if you are being disingenuous or you really don't understand. I'll give you the benefit of the doubt and assume the latter.
BIP101 activates *only* if XT has >75% of hashing power.
No, it activates when >75% of the last 1000 blocks *say* they were mined by XT. That can happen with less than 75% of the hash power saying they are running XT, and you also can't assume that just a block was mined by XT just because it says it was. It is quite possible for me to mine blocks saying I'm willing to accept "big blocks" when I'm actually not.
Only gamblers would run XT marked mining operations without accepting larger blocks. I for one don't believe that a significant amount of hashing power will choose to risk the outcome of a contentious hard fork that they would clearly be responsible for by their deceiptful indication of larger block support. Not a likely scenario IMO.
The only pool producing XT blocks is apparently doing exactly that - just marking them as "XT version" and no more.
See slush's comments about his pool.
But more importantly, no pool in their right mind would be running the XT code yet.
The XT code has been severely lacking in testing compared to normal Core code changes.