Freezing increases in food and accommodation charges might seem like a reasonable thing to suggest, but I can think of several good reasons not to have done so and think the AFPRB have been sensible so far.
Firstly, it would muddy the water between pay and charges. If charges hadn't increased in line with what the rest of the population *theoretically* have to pay (as well as SP in private accommodation) then it would provide the government with an easy opportunity to say "look, we have given the armed forces an effective pay rise by cutting their charges in real terms". As it is, there is no way the government can argue that this is anything other than than a simple pay freeze.
Secondly, not increasing charges to compensate for a pay freeze is of no benefit to those who do not live in Service accommodation or eat in the mess.
Instead I think we should be pushing (admittedly not sure how) for the AFPRB to make recommendations for pay as they always have. If the government genuinely believes our pay and allowances 'compare favourably' to the private sector then they should let the AFPRB make this analysis in an evidence-based and auditable fashion. If the government feels the need to then decline the recommendations of the AFPRB for budgetary reasons then so be it. Personally, I think a temporary pay cap is probably necessary but the government should have the balls to decline the recommendations rather than preventing the AFPRB from doing its job. The AFPRB has said as much in its own report.
If the AFPRB cannot return to making uncapped recommendations next year then I would hope the members resign on the basis that their independence is compromised. Looking as the CVs of the members, I don't think it is too optimistic to think that might happen; none of them seem like 'yes men'.
Whatever happens, there is going to be a major discrepancy between what the AFPRB recommend and what the MoD believes it can afford. Whether that happens next year or gets hidden behind a cap for another few years will be interesting to see.