This is the correct read.
This is also the same as any other software package in existence. In fact, if someone claimed to not use any libraries, I’d be taking a close look at that too.
The key difference is if you’re a paying a company for support and certification of the product as delivered, you can yell at them about it. If you’re using a free product with no support, you can yell at yourself.
Sqlite shouldn’t lock for read, so unless you are writing something at each access, you can have thousands of concurrent reads. The Sqlite website spells this out, and lists its own self as the proof.
This would mean you could not write logs to the database, you’d have to do it the unixy way and put logs in a text file.
Love this, and I want this.
I’d consider some of the secondary requirements, which might then inform what you use for the primary requirement.
Have you ever moderated a site or forum? With user-submitted content, you absolutely need a moderation mechanism, and moderators. This site would be a prime target for astroturfing, trolls, and people with beef. The latter are useful for sourcing content about alternatives, but relying on each contributer to moderate themselves is not going to work.
You’ll also need an effective mechanism for receiving and actioning takedown requests (or telling them to eff off, if that is your wont).
All of that comes down to user metamanagement, and user classes. Simpler wiki software is probably designed for trusted users, or a single user, or might require a plugin to handle these scenarios. Mediawiki is stable and capable, but needs tight site design to avoid it becoming a maze. (Actually, that last bit goes for anything you choose. If you don’t impose a structure, you’re going to get a structure anyway, it just might not be the one you or your users want.)
Extremely old South Park reference?