It's quite likely that there are systemic errors that the experimentalists didn't think of.

This would be an example of negligence on the part of the experimentalist, though perhaps excusable negligence.

Eliezer's point is that an experimentalist might get results contrary to theory (i.e., extremely unlikely, given that the theory is true) even if the experimentalist thought of every possible source of systematic error. After all, extremely unlikely outcomes do happen (albeit rarely, of course). Therefore, the theorist should be able to "defy the data" without implying that the experimentalist made any kind of oversight.