Dear Peter,
They are generally lumped together because they are all 'monodisperse emulsions' or 'monosize crystal emulsions', that is, the crystal size is much more tighly controlled than with the old cubic-crystal emulsions. This inevitably means that they can be finer grained for a given speed (fewer big crystals) but it also means that they have less tolerance for over- or under-exposure (big grain in the former case, less shadow detail in the latter) AND a smaller developer repertoire with once again less tolerance of over- and under-development. The control, I believe, is achieved by 'double-jetting' the emulsion ingredients, that is, adding controlled quantities of A and B simultaneously instead of adding A to an excess of B.
T-grains (tabular grains) are, as their name suggests, big, flat crystals. They're also very thin, apparently enough so to reduce blue sensitivity -- photons can go straight through -- which accounts for the extra dye sensitization in T-grain.
Epitaxial is harder to understand, but as I recalll it essentially consists of growing one sort of crystal on the outside of another. Delta crystals are triangular (surprise!) and correctly exposed Delta emulsions are grainier than correctly exposed T-Grain BUT less grainy than over-exposed T-grain where all the development sites run together.
This is a fairly poor summary, because I'm not an emulsion chemist, and besides, I don't have the time to check all my sources in a reply like this, which of course I would do if I were working on a book. There may well be others on the forum who can correct what I have written. But as you say, there clearly is a difference in look and behaviour from the different monodisperse films: a lot more than there would be if they used essentially the same technology.
Cheers,
Roger