Dr.SurgeonPGH said:

Steady state is faster than people expect: half-life about five days, so you are at plateau in roughly three to four weeks rather than five.

Filing a mild objection. Mild because I might be wrong; an objection because nobody has addressed the case that does not fit. The "tirzepatide is simply better" summary irritates me. It is better on mean weight loss, and the cardiovascular outcome evidence is far thinner than semaglutide's. If the reason for treating is cardiovascular risk rather than weight, the evidence base points the other way.

Adding the numbers, since they settle part of this. Numbers worth memorising for this class: Tmax one to three days for the weekly peptides, terminal half-life about a week for semaglutide and about five days for tirzepatide, steady state at four to five half-lives, subcutaneous bioavailability high enough that site choice is irrelevant.

Worth separating that from the pharmacology, which this thread keeps folding into the same question. They behave differently and the advice does not transfer.

kate.chem said:

The "tirzepatide is simply better" summary irritates me.

Coming at kate.chem’s question from a different direction. The mechanism is more central than most summaries suggest. Receptor agonism in the arcuate nucleus activates POMC neurons and inhibits AgRP/NPY signalling, and the downstream MC4R pathway is the same one disrupted in monogenic obesity — convergent genetic evidence that the target is the right one. Peripherally there is glucose-dependent insulin secretion, glucagon suppression and delayed gastric emptying, but the gastric component largely adapts over months while the central effect persists, which is why the durable effect is appetite rather than fullness.

That is the short version; the long version is somebody else's post.

A narrower follow-up, since the general answer is now clear:

How much of the tirzepatide advantage is the GIP component and how much is simply that the dose ladder goes higher in receptor-occupancy terms?

Closing the loop on my own question.

Update: the eight-week restart pattern people described is exactly what happened. I nearly abandoned it at week five.