Fix s2 photon timing bug

[FaroutYLq]

What is the problem / what does the code in this PR do
Inspired by solution in fuse here, trying to solve the s2 photon timing bug. The detailed mechanism why it fixed things is unclear but we don't care much for a legacy package. Below is before and after the fix:

WFSim needs you:

• Please add a test for this PR, as a bare minimum, make sure it's covered in coveralls!
• Can you add a docsting to all your functions?

Pay attention:

• Due to databases being needed for testing, making a PR from your own fork will typically NOT run the tests. If you then merge master might break

[coveralls]

coverage: 82.78% (+0.1%) from 82.67%
when pulling bb835b3 on fix_s2_photon_timing_bug
into 2317065 on master.

[ramirezdiego]

Funny:

FAILED tests/test_wfsim.py::test_sim_1T - ValueError: Found n_hits less than tight_coincidence
============ 1 failed, 20 passed, 439 warnings in 779.44s (0:12:59) ============
Error: Process completed with exit code 1.

[FaroutYLq]

OK so it somehow passed after running again. I would recommend just merging it then.