Impl `threading` because of I/O bound nature of the script

[storenth]

CPU is fast, but disk 100500x times slower, so we don't want to wait until it respond. The edge case:

cat subs1.txt subs2.txt ... subsN.txt | ./src/pnk/__main__.py

[storenth]

Create large list, perform cPofile, split txt, perform cPofile, file the src/tests/PERF_PROBES.md for v3.0.0 before impl.