I don't think that integrity has much to do with it. I'm sure the guy has always had good intentions. To me, the problem is that what they're doing is just really sophisticated phrenology/craniometry (for the record this is an extreme minority opinion, so very high chance I'm wrong). I commend anyone for reexamining their old results and challenging their own assumptions, because that's a hard thing to do. Maybe he did have nowhere to run, but that's not necessarily a bad thing, especially if the motivation was internal. Maybe he was contacted by some people who couldn't replicate his work, and it caused him to look back at the published body of evidence. That's actually how things are supposed to work, so no harm there.
But I think that saying that maybe the problem is that they only scanned for 5 minutes instead of an hour is incorrect thinking, as it seems that they're misdiagnosing the problem, as if the error is one of degree and not of kind. To me, the entire field needs to reassess (a) what the point of their research is, and (b) whether the approach they've taken to achieving those aims has any bearing on reality. fMRI sure makes for pretty pictures and complicated sounding explanations of areas of the brain "lighting up" when a person does something, but I think it's all really facile and obscures rather than illuminates any truths about behavior.