Elizabeth, I'm not sure what you are referring to when you stated, "After reading the article. my main criticism is they did this with just one component, a relatively low priced device. And that they assumed the not broken in device was not broken in. they admit it had some time already on it) So any extrapolation to the world of all devices is a stretch. But a nice attempt.
ANd to say thee re variations in components. Some change a lot some not at all. So again one test does not offer enough to draw conclusions about all components."
To clarify; I worked clearly with three items; a set of cables, a CD player and an integrated amp. I had duplicates of all three, one set used and given additional burn in, and the others like new. There may have been up to 10 hours of use of the new components. It hardly matters, as whether new or 10 hours, versus the hundreds of hours that supposedly burn in is to change a component or cable, the comparison showed no difference, even when three items were used together. One set was as new, while the other had been used for hundreds of hours and given additional burn in, and used warmed up versus cold start for the new ones.
My conclusion is that was a massive fail for the concept of burn in, and due to the multiplicity of items being compared together more strongly supports a tentative conclusion that no components actually change audibly (excepting the conditions mentioned in the article) over time with use.
Additionally, if you were referencing my article in the second paragraph of your post, in response I clarify that there was not three levels of change; there was no change at all among three pieces being tested together simultaneously. :)