Thank-you @DesperateKingWalk for running this test and sharing the results.
Looks like your tests are giving pretty much the same result as the one published by the CCRL guys after the CCRL 40/15 tests.
In the CCRL-Chess324 tests, instead, SF 16.1 seems stronger than any precedessor. I'm not too familiar with Chess324 and I don't know why in standard chess SF16.1 is not the stronger engine, while this "variant", it (looks like) is.
Would you have any ideas?
The newest version of the Stockfish UCI engine is out.
Official Release and where to download the binaries: https://stockfishchess.org/
To download the source code if you want to compile it yourself for your very own hardware and linking: https://www.abrok.eu/stockfish/
Info on how to compile Stockfish: https://github.com/official-stockfish/Stockfish/wiki/Compiling-from-source