So, in response to a student who had some data ith (i) time varying individual covariates, and (ii) p<1 (meaning, she didn't capture and measure every individual at each occasion), I had a look at these sections in Chapter 11, and realized I'd been negligent in describing Simon Bonner's slick 'trinomial' approach to the problem (which he published in MEE ~5 years ago). A first attempt at correcting that oversight has been added to Chapter 11 (new section 11.6.2). The basic details are in place, but there might be a few bits of language I'll tweak moving forward.
So, kudos to Simon for describing the trinomial approach in the first place, and apologies that I took so long describing it in the book...