Jim L-
Your suggestion regarding ASRS is an action I haven't considered. I'm going to use it and encourage others to do so, as well.
I'll agree there's little use re-inventing the wheel, re:science and duty assignments. I'm ashamed to say, there MUST be a regulatory requirement to drive the application of known science to scheduling, or we won't do it.
As to spatial orientation being a primary cause of night accidents- I'll argue that's only true in the broadest sense. My opinion- The commonest scenario is NO reference and CFIT on departure/approach- the poor slob doesn't even recognize that it's IMC. Also my best guess, the second most common night accident scenario is loss of reference and an attempt to recover with a descent, which is pure and simple suicide through stupidity. Third and fourth are ties- IFR/VFR cruise CFIT. Pick one.
ALL stupid, but stupid's easy when you're mentating as though you'd 3 drinks (or better, that's one result of sleep/circadian disruption); blind; ignorant of conditions; and finally, poorly equipped. Consider, there's NO regulatory requirement for a directable searchlight to do night scenes (that's "off airport" work) in the US...
As to why the industry "splits shifts", flipping from days to night duty mid-week with 24 hours off, it's easy- Look at a calendar. Between days and months, there's one unit of measure- a week. We'd rather have a schedule that "fits" that habitual unit than an odd schedule, starting on varying days. Final factor, some guys can't do 7 nights in a row. Michelle North of the "Wisdom Well" used to preach it, and I've seen that, too.
Fix the problems that'll kill me tonight:
Pilot error- That'll do you, single/twin, IFR/VFR, 121/135/91, day/night, off airport or on;
WX data to my Garmins;
And, NVGs!!! which are coming, but twenty years too slow.