benchmark,lib: add process.hrtime.bigint benchmark
Add a benchmark, and amend the relevant source code comment to state that currently, switching to directly returning a BigInt is not stopped by technical obstacles but rather the fact that using a typed array is actually a bit faster (about 2.5 %, measured locally).
Checklist
-
make -j4 test
(UNIX), orvcbuild test
(Windows) passes -
tests and/or benchmarks are included -
commit message follows commit guidelines