They already do it they put the in library identifier on the games as well as in basket and in wishlist it would be no harder for them to hide it.

I do that basically to ensure I'm not hitting things I don't mean to, at least if I'm understanding what I'm doing, and to remove the whole game rather than just the thing I'm checking for the the in library id. As I said there is probably a better way to do it if you are interested check out how the addons do it.

Hmm. In that case they must have a denormailized local cache of those particulars stored in a (virtual) table. That would mean (if true) that every configurable search criteria would need to be pre-calculated and these new tables created.

I was thinking more of GOG doing it client side you don't get to pick the number of titles that are shown per page sure if you are looking at a particular narrow search or own a tonne of games you might get pages with only a few or no titles but in the grand scheme of GOG web design, meh.

To be clear I don't overly care I actually like seeing the in lib due to GOG failing to properly recognise related SKUs.

I hardly see any difference. But my computer and my internet connection are both quite fast.

For my sins I have studied CJ Date, so I was more interested in a theoretical context; no doubt FLOPS have increased markedly in the many intervening years since I first encountered the relational database theories. That said, mathematically, a join between (large) tables is costly, and its exponential across tables, so the more tables accessed to create the join, the higher the cost. Mostly, though, such bottlenecks can be adequately pre-empted and designed out, given sufficient redundancy in the table-relationship paradigm. It all depends on the choices made when building the construction to house the data and the way they relate to each other.

If, as you are suggesting, computer power has advanced to the degree that multiple agents can bind four-plus tables (of huge capacity), then Kurzweil was definitely correct in his accelerated returns theory! :)