You are right, provided he did not have a notice saying it was MIT licensed elsewhere.
> I’ve contributed to plenty of project that don’t have the per-file copyrights. It’s a choice not a mistake.
I would consider it to be both a choice and a mistake. The two are not mutually exclusive. There is no evidence in the fork that he is the copyright holder of the original code and it looks like Microsoft is. Part of that is Microsoft’s fault, but part of that is the original author’s fault for not including per file copyright notices, such that Microsoft could add theirs and be the sole one listed in every file.
I would not be surprised if Microsoft’s legal department doing a scan of public repositories for stolen code mistook him for infringing on “their code” given that they have no information that he authored it rather than their employee. It sounds absurd, but it has happened. I know for a fact the sg3 utils author added copyright notices to his code examples because he was getting contacted by companies, whose engineers incorporated his code into their projects without attribution, that thought he had stolen their code:
https://github.com/doug-gilbert/sg3_utils
I know that because he told me by email in 2013.