Just because a single page on the internet used it does not mean the term was used in the industry. It also doesn't matter if it technically makes sense, although it's a stretch (because there is a server process, you just share it).
It's a marketing term, and a poor one at that. That's why SQLite even describes itself as a embedded, in-process database without client/server architecture. Because that's the common jargon.
I'm surprised at the endless defense of a marketing buzzword and the argument over which marketing definition is the "real" one. This is bikeshedding at its finest.
Nobody was saying it was a buzzword in the industry back then (actually...[1]). It was being used as a normal English word, not some misleading marketing drivel like the new use of the term. Somebody who has no shirt is shirtless. An MRE heater that doesn't catch fire is flameless. Somebody who isn't witty is witless. Somebody who doesn't have a clue is clueless.
Are you starting to notice the pattern here? "less" is a suffix that can be applied to nearly any noun to describe the trait of lacking that noun. When this is done, the meaning is clear to native English speakers who've never heard that combination before. When this article was written in 2007, the meaning was clear. SQLite didn't muddy anything, didn't redefine anything.
[1] The term 'serverless' was in fact in use in the decade and a half prior to 2007: https://books.google.com/ngrams/graph?content=serverless&yea... I don't know if the term was being used primarily in a technical context, but I have little doubt the term was being used to describe something that did not have a server. In fact I'm just about certain that spike was the term being used in a tech context. I've found you an example of the term being used in 1995 (DOI 10.1145/224057.224066):
> A serverless network file system distributes storage, cache, and control over cooperating workstations. This approach contrasts with traditional file systems such as Netware, NFS, [...] where a central server machine provides all file system services.
This is not precisely the same way that SQLite has used the term. Rather, they're using the term in an intuitive natural way that fits their context, just as SQLite used it in 2007.
Here is an example from 2007 (DOI: 10.1109/TNET.2006.886289):
> Abstract—We explore exploits possible for cheating in real-time, multiplayer games for both client-server and serverless architectures.
Here we have the term "serverless" clearly being contrasted with "client-server", which seems very similar to the way in which that SQLite document used it.
It's a marketing term, and a poor one at that. That's why SQLite even describes itself as a embedded, in-process database without client/server architecture. Because that's the common jargon.
I'm surprised at the endless defense of a marketing buzzword and the argument over which marketing definition is the "real" one. This is bikeshedding at its finest.