It may be he realized that his analysis was poor given the comments and hence didn't continue. I assume he was encouraged to write something backing PACE.
He is mates with the Goldacre/Wessely crew, but I wouldn't assume that they put him up to anything.
I think that he's part of a group that share values I consider rather disrepuatable, and according to their values, PACE is not a particularly bad piece of work. It didn't fabircate data. It was relatively large. It (roughly) followed most of the key guidelines on how research should be conducted, and while it deviated from its protocol there was still a significant difference for their pre-specified primary outcome, and claims about recovery are a relatively minor secondary finding. The Lancet paper (compared to the promoting in the press) was relatively cautiously phrased.
He acknowledges that PACE cannot show CBT/GET are any more effective than placebo, or that the changes in questionnare scores represent any real improvement in health, but that doesn't particularly concern him. I think that he's much more concerned about the way science operates according to the standards set by researchers for themselves, rather than whether patients are being provided with accurate information about the efficacy of the treatments available to them.
In a lot of his blogs there is a lack of moral concern about the impact of poor science. I had thought that this was him playing the game, and being restrained in order to prevent people from dismissing his criticisms as too extreme or emotional, but some things I read from him makes me think that he just isn't outraged by the low standards many researchers set themselves, and he's said that in his day-job he himself does some of the questionable things he complains about in his blogs. Maybe this was him attempting to be disarmingly self-depreciating, but I think it could well be that he just doesn't see poor science as immoral in the way that many patients do.
PS: None of that is to ay that personal connections, a desire to avoid pissing off powerful instituations, etc, don't also play a role.
PPS: Also, his blog wasn't pure hatchett job. He could have been much more brutal in response to Julie if he'd wanted, and in a lot of ways it seemed that, as well as not being terribly concerned about the moral issue of patients being misled, it was a bit lazy. I got the impression he'd really focussed on the data Matthees had secured, and hadn't taken the time to look at a lot of the wider issues with PACE.