Something needs to be done to improve Google Drive scanning. It’s far too aggressive and definitely not optimized for a cloud service provider. You can’t treat a cloud service like a regular filesystem; it’s simply not the same. I’ve commented on this (here Improve performance of Google Drive scanning and here Option to sync ALL metadata and libraries to iCloud).
I’ve been banned for the last three days (one ban after another), and I realize now it’s likely due to this part of a scan (https://firecore.com/sites/default/files/upload/img_7349.png), because most of the media has been found (which also gets me banned just to get acknowledged), so there’s always this other part that “Fetch[es] thumbnails from Google Drive,” which is even worse for bans. By the way, the numbers in that image reflect what is being fetched after the third ban, so there are thousands of other items that have been retrieved so are no longer counted on the right side of the “### of #####.” I would probably lose my account if I attempted to let it continue through the rest of the items, as I’ve already been hit by one every day for the last three days. I can only assume this means Infuse is downloading each item and generating preview thumbnails out of them, because I do not store any metadata alongside my actual data. There needs to be an option to not automatically do this (and not automatically start scanning the drive, either). Every time I open Infuse, it just starts this process again and there’s no way to stop it. I can’t have it open for more than an hour before the ban. The only reason I would even have the app open longer than that is because I’m attempting to fix incorrect metadata matches. At this point, I’m reluctant to even open the app at all, because I might forget and then get banned again.
Ultimately, as I’ve mentioned before, this integration to Google Drive just really feels like it was an afterthought, with the existing filesystem scanning just pointed to the cloud service, which is not how to scan cloud service directory structures; they have APIs for this, and specifically to retrieve only changes, which both Plex and the open source tool, plexdrive, have each implemented successfully.