The people doing the reanalysis used the PACE investigators' criteria from their own pre-published protocol. And this is the first time I've heard it claimed that PACE was too small to find an effect for these therapies
There is an argument that given the small effects the trial was under powered to show anything in terms of recovery. But that is bad trial design it doesn't mean that you can substitute in an alternative 'recovery' definition. It simply means you cannot conclude anything from the trial. We should remember that the power calculations for the trial were done expecting a much bigger effect than actually happened. Of course having large numbers does tend to allow significant tests to succeed.
Looking at the data it is clear the quality of recovery is better for those in the SMC or APT group especially with the 6mwt.