Tests were done using different hardware.
What version of LiteSpeed was used, was it the latest version of LiteSpeed?
It appears that the servers were into two different locations. The Nginx one was closer (observe the connection times)
Why were document lengths different?
To conclude, it looks like a pretty poor comparison to me, however - I really suggest you yourself do the benchmarking - each environment and setup is unique, and there's no telling what is better for you until your environment has been tested with both; so yes, same php.ini, flags, etc.