Of course Simpson's is heavily at play, of course especially when lumping <25s with >75s. Not sure I see how one can get from the (UK) data to statements like 'more than 99% of recent deaths were among the unvaccinated', which seems off by at least an order of magnitude.