This Ancient Language Has the Only Grammar Based Entirely on the Human Body

Eh, it's misleading, but not entirely incorrect.

You're correct that the -ed suffix does originate in the Proto-Germanic preterite, and there it was an innovation. The vocalic alternation in strong verbs is easily derivable from regular ablaut in Proto-Indo-European perfect/stative verbs (Ex. Greek pres. λείπω, perf. λέλοιπᾰ). The weak suffix has no counterpart in the rest of Indo-European, but especially if you examine the East Germanic data from Gothic, where the plural forms are more complete, that the suffix likely comes from a periphrastic construction with the past tense of PGmc *dōną is pretty hard to avoid.

For instance, the Gothic equivalent of 'to live' is 'libān'. The plural past forms are as so:

1pl: libaidēdum
2pl: libaidēduþ
3pl: libaidēdun

Now, in PIE, we know there was a reduplicated verb *dʰédʰeh₁ti, meaning 'to do/make'. In fact, it directly survived in Greek as τῐ́θημ. If we take that verb into Proto-Germanic, these are the expected plural forms:

1pl: *dēdum
2pl: *dēdud
3pl: *dēdun

Given that these essentially match the forms of the weak past suffix in Gothic perfectly, it's hard to draw any other conclusion. At the least, the -ed suffix is almost certainly related to 'did', and quite probably is a descendant of the exact same word as the Proto-Germanic equivalent of 'did'. That said, "did use became used" is not an accurate account of what happened (to say nothing of the fact that this happened long before the verb 'use' entered Germanic).

It is a bit of an interesting question whether English 'did' is a direct descendant of PIE *dʰédʰeh₁ti or if it was re-formed by appending the then-standard weak past suffix on to 'do', though I'm not knowledgeable enough to definitively say. PGmc *dōną didn't actually survive as an independent verb in Gothic or Old Norse, and so by the time we see it for the first time in Old English, the weak past suffix has already long been established. That said, in Old English, the attested form is dyde rather than expected *dede, so something weird happened at the least, though that's hardly uncommon with extremely common verbs like this.

According to The Online Etymology Dictionary:

Old English dyde, past tense of do (v.). The only remainder in Germanic of the old linguistic pattern of forming a past tense by reduplication of the stem of the present tense. Far back in Germanic the equivalent of did was used as a suffix to make the past tenses of other verbs, hence the English -ed suffix (Old English -de).

It's interesting that a similar thing didn't happen for the word shall or will, in order to create a future tense. For example, using a suffixe from "will", like "-ll" or "-ill" for a future tense.