I would counter it was that, and your history doesn't conflict with mine - it explains it.
It's true that MINIX was held back by the fact it had to be distributed as "Core system (from book) plus third party ports as patchsets", but that doesn't pertain as to why Torvalds wrote Linux, except in explaining why there was no ix386 version of MINIX to begin with.
The MINIX community maintained standard ports of MINIX to each architecture. To install MINIX you'd get the book, combine it with the port, and you'd have your system. Your WANG drive patch would have belonged in one of the ports, not in the core system (frustrating if the patch needed architectural changes as that meant there was no practical way to distribute it.)
It wasn't a particularly effective way of maintaining an operating system, and MINIX suffered as a result. And one way it suffered was in not having a standard ix386 kernel. Why Torvalds didn't port the MINIX kernel himself is something only he can answer - I'm sure there are valid technical reasons, and he may also have been frustrated with the community core+ports model, but I like to think he also wanted to scratch an itch, to figure out how a kernel worked and write one himself. And I'm glad he did.