The idea of break-in has been studied. Here is Dr. Toole on it:
I saw the measurements in a private meeting at Harman. Yes, the driver itself shows noticeable change. But put it in a box that the difference shrinks to almost nothing. As noted above, listening tests were performed to confirm the same.
As to his car analogies, if you take a bolt and heat and cool it, it will show a difference in length. Put it in an engine fastening something, and you won't see that result. He talks about suspension of a car. How come that car doesn't ride softer after "break in?"
If the difference is real, then he should build two identical speakers, test them blind with people listening to music and confirm they can't tell the difference. Then break one in and test again. If the results are positive this time, then we know there can be a difference. He makes speakers so he should have no trouble doing this test.
Now, tell people there is an effect, and they will hear it. They run the speakers for N number of hours, then listen more attentively and they will hear more detail, air, etc. and consider the speaker more broken in. In reality the sound hardly changed, what changed was how they were analyzing sound. That, we know changes a ton.