Medieval clothing database

I have over sixteen thousand pictures of medieval paintings, frescoes, illuminated manuscripts, sculptures and extant garments relevant for the research of medieval clothing saved on my computer. They were found all over the web; most of them have at least partial description, but some have such imaginative titles as "sfdnjnjnuu" or "654484848498" and are not sourced at all. Many are copies of the same artwork in different qualities, or even the exact same image saved under a different title. The folders are a mess, really, but I am sourcing and renaming the images properly, albeit the progress is slow.

Many, if not most of them, are from wikipedia, WGA or various online manuscript libraries; many are screenshots of the image of the artwork if the hosting webpage did not allow for the images to be downloaded in good definition. I'm looking at you, Cranach archive, and you, National Gallery. It was no fun at all making screenshots of pieces of a painting and pasting them into Paint, matching them pixel by pixel, just to have a decent resolution of the artwork. (I am sure there is an easier way to do it, via html script or something, but that is far beyond my extremely meagre programming abilities.) You can see how that effort turned out here and here, for example. Btw, this is why I love Met. They are doing it exactly right, just as museums - institutions focused (among other things) on preserving and spreading the word about the cultural heritage of mankind - should. That is, being user-friendly and accommodating in sharing high definition images of old artwork, putting them into public domain.
Anyway, I strayed away from my point. I've been wondering if there is a way to make a public archive of all this art - put together, properly sourced and attributed and documented, from all over the web, in the best definition available, with a good system of tags. I tried something like this with illumanu, but tumblr is entirely unsuited for databases of this kind. To find a functioning, efficient way to tag the images - including the dating, area of origin, artists or workshops, current location, content or title and also, importantly, all the depicted objects relevant to the database's purpose - proved to be quite a challenge, especially because the attributions are often dubious and the clothing terminology very often inconsistent, too broad or too misleading or even non-existent for a particular feature. Imareal is not so bad an example of how the tagging system could look like, but the database is otherwise fairly confusing, hard to search and its links are often broken.
So, I guess for now I should busy myself with clearing out my folders, but it would be such a nice thing to be able to be handed over images of a specific item of interest from artwork from all over the Europe (and beyond!), without having to do an inefficient search all over the internet or zooming in the altarpieces on Google Art Project and hoping to find the one detail you just need for your research.