Burning-in as a non-linear experience

I know there has been loads of discussions on the burning-in process of components as well as their parts. If someone does not believe it happens, please do not respond. This is to address mainly the experience people had in burning in components or their parts. The burning-in process is usually described as a linear process, getting from bad to good. But in my experience, and in my system, it is not a linear process. It usually starts from pretty good, to better, then worse, then better, then horrible, and finally wonderful. I was wondering if people had a similar experience. These are pretty drastic changes... And this topic is mainly to focus on this particular thing: non-linear changes during the burning-in process. I had this experience with Furutech NCF outlets, but now I can see the same may apply also to my new capacitors. I have recently replaced my Samsung capacitors with Nichicon LGL2G821MELC40, not an audiophile sort, but the only ones that I could fit in. People are reporting about audiophile capacitors needing a lot of time burning in, I was wondering also if non-audiophile capacities might sometimes need the extended time in a high-resolution system and if their burning-in might also happen not exactly from moving gradually from bad to good.