An additional thought to add to this direction of seaward flight speculation.

Our prevailing winds are west to east bring moist marine air inland. Surly
there must be a detectable gradient of "marine smells" for want of another
term.
Even we humans detect ocean odors long before we sight the water. Also
desert species are credited with being able to "smell" water and direct
their paths to water holes.

The directional orientation mentioned above ( see my video comment above ),
the odor gradient, residual light suggestion and other yet unknown
behavioral controllers all require basic physiological studies and marble
murrelet investigations have not progressed much beyond uncontrolled
habitat/behavioral studies. There is a next level of research to explain
the reason why questions.

Bill