[DaveBlake]

If someone can scope out the issue - repeatable problems, features needed etc. , then that could help me when I get to look at it.

I can do that, no problem. Maybe even submit a patch.

Great, what can I do to suck you in

Having something that you want to fix is a great way to find out about how Kodi works (or doesn't), it focuses the mind. There is far more work to do than devs, so contributions from those that have the skills to do so are always welcome. But the review/approveal process can be slow, again not enough devs with time and energy, and hasty changes can easily upset Kodi - it is a chicken and egg situation. I know it gets frustrating for all concerned. I would do my best to support other's efforts, and gladly collaborate.

I haven't looked at any of the code in over a year, and even then it was just cursory peek, so if you could give me some suggested starting points for peeking through the code, that'd be helpful, Dave.

Seems that you already found the area to start in yourself. I really am vague on thumbs handling, I have never traced it through as too busy with other stuff. It is really hard not to become spread so thinly that I become ineffective.

If you want to keep talking aloud, or have specific questions I will try to engage. I generally know the scanning / scraping process, and the music database, just not the images/texture side which is a little different. I have always just assumed that what was there, when I tripped over it, was right.

Set up IDE, make a Kodi branch and if you post stuff to Git I will look at it there and discuss too. If nothing more then summerising the curent design is useful.