Well I'd rather it be like this than the opposite extreme where companies take any (MIT licensed) software they want but never give back or publish FOSS of their own
* taking the name of an MIT licensed thing which is entirely not covered by the License and I’d at least consider that bad form.
* A breach of the MIT license by the Facebook engineer - specifically removing the license and passing it off as their own work.
So the author is not - as you seem to imply - complaining about someone using their software under the MIT license.
How different does one project's code have to be from another similar project before you can claim copyright and slap your own license on it?
If a dev simply copied the entirety of someone's code and simply renamed some variables, could they claim it's their own original work (without worry of legal recourse)? I'd figure not, in which case, how much of a difference is different enough?
That's why reimplementations of license-encumbered software have to be careful to have one team study the software that is to be reimplemented and describe what it does and another teams reads this specification and implements the substitute without ever looking at the original software.
This would most definitely not be enough. That would basically make all licenses useless - commercial and noncommercial alike. The MIT license is no different than any other license in that regard.
In case of the MIT license, whenever I copy any code (files or parts), I always place a comment right next to the section, pointing to the source of the copy and containing the license of the code/copyright marker. Honestly, what’s the cost? My ego doesn’t take a hit from admitting that I found and adapted something useful. The possible uses of the code are unaffected. License situation is clarified for audits. The original author gets credit for their work. There’s only upsides.