I like the idea, but (at least at the moment) don't want to spend the time to implement this. And I don't really like the idea of having to connect to an external server 50,000 times per week, but I also don't like to install unknown software.

I'm going to rewrite the basics of my scraper (it's my typical addition upon addition structure), and after that sha256sum each archive to one file per directory (up to 10,000 files). Then I'll leave it to the viewer to archive those checksums at any moment.
I realize someone could claim I can use Apache2 logs to see if anyone has accessed that file, and thus know whether or not I can tamper with data without evidence. If anyone doesn't trust me with this, feel free to archive those files as often as you want.
On second thought: I could use OpenTimestamp on those checksum-files, which would only need to be done 5 times a week. I'll see if I have time