Investigate the breakdown of the `sleepy_boi` test

[workingjubilee]

Happened in #159. There's no obvious sensible reason why it changed. I should not have had to comment out that assert from snooze().

[eeeebbbbrrrr]

Is postgrestd using one of Postgres' functions for the time? I don't recall which function it comes from, but its SQL now() function is fixed at transaction start. I think the SQL timeofday() actually ticks.

[workingjubilee]

What's even weirder now is that it seems to even run the assert if it is commented out...? #167 will kill it off. It's possible something is happening due to syn altering the line breaks or formatting. I should investigate this more closely sometime soon.