This stuff is of Indo-European (and then Germanic) origin and as such it had more grammatical cases and genders (even as recently as in Old English). In this sense English must have lost some entries through neutralization of those grammatical categories in some forms.

It's not just lexemes and forms that may get lost. For example, "who" - "he"- "the" could be understood as reflections of corresponding words. Neither their semantic nor formal correspondence is perfect today. Did English lose any of those words? Nope. Did it lose the correspondence between them? I think it did.

Very interesting - of course, I was mainly referring to stems rather than derived forms. Though am I right in thinking that it would also be "whom" "him" "them"?

On that matter, it's also interesting to me that "he" is actually a third-person pronoun in modern English, rather than something like 'this person' as would be expected from extrapolating from the others' patterns. Has it always been this way?

I'm afraid I'm unable to answer the main question as asked, primarily because I haven't gone through the exercise of looking up the etymologies and Proto-Germanic, Latin, Greek, Sanskrit... parallels of enough candidate words for a clear picture to emerge. Looking at only the modern forms can be misleading because the forms not only diverge, but sometimes also converge ("analogy"). "him" and "them" seem to have separate histories, at least partly. "him" was the Old English dative of "he", but also of "it". "they"/"them" is from Old Norse. The "m" in "them" , taken as a dative marker, is pangermanic, but the corresponding lexeme in Old English had a different vowel than Old Norse and yet different than what English "inherited" (or "borrowed"?) from Old Norse.

Regarding "who" - "he" - "the", I'm afraid that I was oversimplifying to the point of not being quite right on that point. Old English had a word "sē" which was both a definite article and a pronoun (= article usable without the noun). Masculine singular in both cases; Old English had separate articles by gender. In late Old English this word changed to "þē", the predecessor of "the". Whereas "he" comes from "hē". I'm afraid that when I label the modern "he" as "proximal" and "the" as "medial", that this is just in my eyes, influenced by the form more than by the meaning. However, I do think that "who" (OE "hwā") and this predecessor of "the" are related. I was trying to exemplify that sometimes an item in the paradigm isn't lost altogether, but an item can somehow move off the radar, formally or semantically. And other times "analogy" or "borrowing" could hypothetically create a perfectly fitting item in a cell where there was a gap a few hundred years ago.

I see, that makes sense. I was somewhat under the impression that since all of the forms that I found were basically unchanged in spelling except for the initials, that this would be the case for the others as well, though I see now that my impression was flawed.

As it is, it does seem to be the case that "he" has drifted; etymonline describes it as coming from PIE *ki-/*ko (this, here), while "the" either ultimately comes from either PIE *so (animate demonstrative) (the precursor to OE se, masc. demonstrative) or *to/*tod (inanimate demonstrative) (precursor to OE þæt, neuter demonstrative) after the collapse of gender.

More generally, the pattern seems to be *kw- (interrogative), *k- (proximal, this), and *s-/*t- (medial, that, animate and inanimate respectively) e.g

where: from *kwo
here: from *ki + r
there: from *to + r