Adding the clinical framing, because it changes how the question reads. The distinction that resolves most of these threads is between what is true on average and what is true for one person. Both are real; they answer different questions and get quoted as if they were the same one.

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

Reporting back.

For the archive: it took about six weeks rather than the two I expected, and the difference between those two numbers was the whole of my problem.

andrew_nyc said:

The distinction that resolves most of these threads is between what is true on average and what is true for one person.

Adding the part of the answer the thread has not reached. The version of this that has an answer is narrower than the version being asked. Narrow it and it becomes tractable; leave it broad and the thread will produce nine confident and incompatible replies.

Adding the numbers, since they settle part of this. Give anything pharmacological four weeks before you judge it, and give anything measured weekly a four-point rolling average before you call it a trend.

Happy to go further on any of that.

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

What would you measure differently if you were starting again?