abouchez / mormot: huge performance boost

[synopse]

Since crc32c is a perfect hash function for our dataset, we don't need to store the names within the hash tables.

• The performance is much higher;
• Code is even shorter;
• I kept the "old" code as reference in the folder (with another project name);
• Please try with 16, 24 and 32 threads: on my PC, I was able to use ALL my threads with this new version: it is likely that on the benchmarking hardware, we could be able to use the 32 threads or the Ryzen CPU with benefits.

👍

[gcarreno]

Better performance is always good!!