All their tests are CPU-bound, not IO-bound, namely -- archiving, rendering, encoding video. Performance degradation happens with transactional, IO-bound tasks (think noSQL databases, ad serving, trading)
It is possible Microsoft has mitigated the issue in a way that has much lesser performance impact. Maybe they had a highly tuned feature to enable kernel page separation already coded but disabled. I won't be surprised if even the Linux implementation is tuned to the absolute limit in the coming months.
But if you scroll down to "Windows-Benchmarks: Anwendungen" you can see that most applications do not have any performance hit with the Windows patch.
Only M.2 SSD seem to be affected.