So if this is true, the "independent" review of the data is yet engaged, with a group the PACE team agreed to share the data with. Not a good sign, they know enough people that would be ready to happily do the job. And there are several countries where that could take place. (My bet is on the Netherlands.)
Yes, it sounds like they're getting their cronies to "independently" re-analyse the data so they can point to the re-analysis and say there is no need to share the raw data with skeptics. Extremely worrying development.
As for Gelman's blog post, inexplicable. Surely the issue of normal range is the key issue when you have a continuous outcome variable like SF-36-PF. After all, what difference does it make if scores went up by (statistically significant) 10 points when mean score at the end of treatment is still in the range you see in heart failure and lupus.