I think this is mostly specific to the T-libs review queue. The 2 PRs you linked have assigned reviewers with 98 and 94 reviews currently assigned to them, despite both actively doing reviews. In general it's been the case for years that the demand for library reviews has been higher than supply of people willing to do library reviews on volunteer time with any sort of consistency – not for lack of people trying it, and then quickly quitting. As far as I know currently nobody is paid to do these reviews because (ᴀ) companies don't tend to see this as high enough leverage (in contrast to compiler work, for example, or in contrast to sending library PRs on paid time which obviously exacerbates things) and (ʙ) even the one reviewer who formerly had a paid arrangement burnt themself out over it.