TLONM became an sort of "instant classic" in digital media circles mostly because there was very little written on the subject and because TLONM came out of MIT press. Just last week, I saw TLONM referenced in a Swedish article written for digital archivists.
Unfortunately, people in digital media circles don't usually have a background in computer technology or in any kind of formalized theory/philosophy, so they're easily hoodwinked by misstatements about the nature of computers, databases, file formats, etc. Manovich uses misstatements of this kind to make his "theory of new media" sound plausible, while ignoring critical issues with the examples on which he bases his reasoning.
I don't buy in to the (usually) right-wing claim that academics are trying to trick people with overly complicated language; different disciplines require their own terminology and the ability to take knowledge of the terminology for granted when writing. But Manovich is the rare case where a writer does actually use language as a way of obfuscating their own faults. TLONM is a slog; it's dull, the language is needlessly obtuse, and many of the facts are wrong. I say this as a person with degrees in both computer science and art.
I read his book over 10 years ago and I'm still annoyed every time someone references TLONM because it means they have been misinformed by Manovich and unfortunately don't have the background/tools to understand why.