>Now, if faciat is a present subjunctive by virtue of
> its being obligatorily combined with a non-past main verb, then
> faceret is the corresponding past.
Only in these constructions. In other constructions, it is the "imperfect"
subjunctive which carries a present meaning, while the "present" subjunctive
carries a future meaning. So we can't read a past temporal meaning into the
imperfect. (There is in fact only one construction where the independent
imperfect subjunctive can refer to the past.)
> [reference to redundancy in language]
> I truly and honestly can't see what is wrong
> about that.
Nothing wrong with redundancy - all languages do it. But your argument has
collapsed. The fact that form A is regular after a past verb, and form B
after a non-past verb does not mean there is redundancy. It does not
necessarily mean that Form A is also marked for past. It may be or it may
not be. In the case of the Latin imperfect subjunctive, there is nothing to
make us think it is, and an lot to make us think it isn't.
Peter