Schools always have difficulty getting the offer/acceptance ratio correct. They know how many places they want to fill but they do not know how many offers they need to make to get there. Their best guess is to go by experience, looking at previous years. This means a combination of over-offering and a waiting list as an insurance. This means in turn that there will be some candidates (5? 10? 15? nobody knows) that will be interviewed who are at the very bottom of the waiting list and have no realistic chance. So I don't think what some parents say is plausible, namely that a candidate might be at the bottom of the academic exams (barely making the cut), but improve their ranking through the interview. I think that's unlikely. These candidates are not interviewed to be tested, but because of this reason: if the worst were to happen and many more families than previous years turn down places, then the school might have to reach very deep in the waiting list. For it would look bad on a school to make an offer to a candidate that it had rejected in writing a few weeks ago. So the waiting list must always be longer than what the school expects, which in turn means some candidates had no realistic chance when they were interviewed to begin with (that's how insurance works).
Of course, the above analysis only applies to schools where the second round is just interview. But it explains two aspects of the admissions process in some of these schools: (a) some interviews are not at all academic (b) some candidates on the waiting lists never get offered a place.
Now, I think all this is extra-complicated this year, with the uncertainty caused by the introduction of VAT. There is a danger that more parents will choose state schools in the end compared to previous years, having applied to indies as an insurance. Nobody knows, of course. But it seems to me that, if I was running admissions there would be two choices:
A) Keep the number of offers fixed (which is always more than the places) and call back more students than previous years so that (if the worst were to happen) there is a longer waiting list to go to. But for the waiting list to be longer, a school would have to interview more pupils. And some of those in that list, if acceptance rates drop, will have a realistic chance whereas they wouldn't last year.
B) Alternatively, a school can just make more offers to being with, anticipating some drop in acceptances, and keep the number of interviewees fixed. This second approach is more risky, however, as it could result in more acceptances, forcing the school to create another form at short notice.
Given that (A) is more risk-averse than (B) I would venture a guess that schools are interviewing more pupils this year. Or they might try to combine them (slightly more interviewees, slightly more offers than previous years). Of course, there is option (C), to do exactly as a school has done in previous years. But I doubt schools, even the very top ones, will be thinking in that way. Institutions usually plan for the worse.