The argument I have presented has several parts, so it might be worth brining them together in a summary.
Where others have accepted the framing of the opening post, and argued about presentism, eternalism, memory traces, causal inheritance, the block universe and so on, I want to question that framing, and I hope, dissolve it - to show a way out.
The first step was to show that the past is poorly treated, if it is understood as something to be quantified. Looking at formal treatments of temporal logic shows that the question “does the past exist” is malformed; it’s a grammatical act that gives the illusion of there being a something - like the little man who wasn’t there.
Next, we can see how this might be used to reframe Peirce’s claims. He puts his account in terms of truth, when they might better and less controversially be put in terms of belief; he talks in terms of ontology when the issue is epistemic. The usefulness of past-tense statements might well be addressed in terms of their future utility; but their usefulness is a very different issue to their truth.
The third point: the root error here is one of equating belief and truth; of thinking that all there is to the truth of a statement is an attitude towards that statement. But the truth of a statement is a different kind of thing to the attitude we adopt towards it, that being believing it, knowing it, asserting it, doubting it, or whatever.
The final part of the argument is the observation that the T-sentence, “p” is true ≡ p, provides a minimal account of what it is for a statement to be true, without introducing ontological implications. The ontological implications might well be there, but they enter via “p”, not via the T-sentence.
Now this argument has the result of dissolving most of the context of the thread, but leaves us with most of Peirce’s conclusion intact. What folk believe about the past will drive what they do in the future. But it remains that those beliefs can be false.