Estimated remaining download time using e.g. the exponential moving average, and displayed in a format of "A days B hours C minutes D seconds" (or "A d BB:CC:DD) would be nice, just to give some rough estimation how long it can still take.

https://stackoverflow.com/questions/2779600/how-to-estimate-download-time-remaining-accurately/44577776

I guess update and verify could have a similar estimation as they also can take a long time, but maybe they'd be very inaccurate if it was based only on the number of remaining game titles (still to be updated or verified) as the number and size of files can vary so much between different titles... I presume verifying e.g. Akalabeth takes much less time than Cyberpunk, but the estimation would possibly consider them as equals, unless it was more detailed like based on the number of files (not titles) or even the size of remaining files...

I can't really tell which I'd prefer. Subfolders could tidy up the look (e.g. moving the various separate patches to a separate subfolder), but then in some cases they might just make it more complicated. If I have to decide now, keep everything in a single folder.

You can get a pretty reliable guess at how long Windows / Extra files take to download (pretty consistently around 40 GB/ hour but can dip below that occasionally ), but the Mac / Linux servers seem to be limited and much less consistent ( they are usually ~15 GB/hour for me but they can dip far far far below that and do fairly often ) so I don't know how useful that would be overall.

You're correct that games take hugely different times to verify: Baldur's Gate III or Cyberpunk or similar games are a bit over 100 x the size of the entire Ultima series.

Would have to do something like a 2 pass system there, ie scan for everything that doesn't need reverifying, make lists of stuff that can be MD5 verified, stuff that can be zip checked and stuff that can only be size checked and then do each list individually by total file size (it's not a perfect estimate (several smaller files have more more overhead than 1 larger file ) ).

I specifically covered that part:
===================================================================================================
Nested folders would work for me, but the reason I did not nest them in my suggestion was that if someone's only ever downloading English Windows, for example, they will have a pointless double click. Everything will be Window\English, making that subfolder feel kinda pointless.

Still, better that one pointless folder in such a scenario, than everything, including goodies, being in one place.

And speaking of pointless subfolders, I double checked the image folders. They are:
!images\bg_url\images-2.gog-statics.com
and
!images\image_url\images-3.gog-statics.com

What is the purpose of these "images-2.gog-statics.com" and "images-3.gog-statics.com" folders? Unless there's a very specific reason that I am yet to discover, I think they could be ditched. Just put the images into "bg_url" and "image_url".

There is indeed a very specific reason,, mimicing the URI structure lets me detect when the images have been updated cheaply and automates the cleanup after the new ones are downloaded. .

Doesn't it mostly depend on the downloader's download speed? It varies a lot, depending both of your internet connection, how congested it is, or in which continent you are (e.g. lots of Australians seem to have download speed issues with GOG).

Yeah so for verification the "estimated time" is much more just an educated guess, but could give some rough idea I guess, making the (wrong?) assumption that on the average the remaining game titles will be of similar size as the already verified games. So if it took on the average e.g. 7 seconds per game title to verify the files (while in reality it keeps changing wildly, big games taking 50 seconds to verify and smallest games 2 seconds, for example), it makes the bold assumption that this average will hold true also for the rest of the games (while in reality it probably doesn't).

Still, could give a rough estimation whether running a full verification will still take 8 hours or three days to complete, I guess (with a very large library). And the estimation would be relevant only if you are doing a full verification over all your games, not just those which have recently changed or have been added as new.

Anyway, I feel the estimated time for downloads would at least make sense because it is more feasible and more accurate to estimate.

Oh, fair enough.

To elaborate more on the second request:
Asterigos: Curse of the Stars is a good example.
Main Game
Expansion

So it would be nice if GOGREPOC was able to grab both of them. Now, of course, the downside is that other DLCs use the same image as the main game, so the risk is that we may create several pointless duplicates. So the options are:
A) Accept it, since images take up little space
B) Compare sizes and / or MD5 checksums of downloaded images and skip/delete the duplicates, if they match. The non-expansion-DLC images for Asterigos use different URLs and file names, but the images are identical once downloaded.