I think we should try and deprecate *readtable-alist* but keep it
i'd like to speak up against deprecation.
my reason: in-readtable gives a high resolution for reader control, but i think such a high resolution is often not needed. e.g. i'm perfectly happy with our current setup where the package also defines the readtable, and thus one doesn't need to copy-paste an in-readtable form into every file.
such a setup would be impossible without *readtable-alist* or an equivalent infrastructure.