I had indeed dropped it as a serious suggestion, and much prefer my final post before Chad’s. But Chad brought up the point about this being impossible in Plato.
Again, just talking about aspirate assimilation in the hypothetical, and not as a serious suggestion for John 20:18, I completely agree with that it could not (and did not) survive in transmission. However, a ταῦτ’ ἅ in an original or early version strikes me as exactly the sort of likely error that could have been misunderstood in this sort of work at the earliest date. And exactly the sort of error that would have been made by the author. Allen and Horrocks make clear that the sound may have stopped being pronounced in many areas at around the time this Gospel was first composed or written down (the very early second century AD). The author, simply from internal evidence, was not a highly literate individual.
The inconsistency you point out in BGU 846 was used by Horrocks to make the point, hence “erratic”. The writer no longer had the sound, and so dropped or added it at random in his text.
I’ll give the Grammatik a read. Everything is grist I’m afraid. In a few decades maybe I hope to have something intelligent to say.