So basicly its 99% the lisence that drives this project am I right?

I mean yes its nice to have some cool features and stuff, but if you really dont care about speed why the hell would you wann use C? Then you would use python/javascript or something like that. But for some parts speed matters, and then in most cases 10% more speed or less matters also. So if thats right and the benchmarks are true that in 99% of the cases this new compiler results in noticable slower blobs, I dont see much usecases.

Or is it guarantied that you can take code that runs with this compiler and compile it for production systems than without any code changes with gcc? Than you could maybe use that for having better developer-experience and in the end compile it with gcc so the end users get the speed on this parts that are written in C to have speed. (The only reason a sane person would write code in C).