Zork Grand Inquisitor: Fix Determinism Issues on Fixed Seeds

[nbrochu]

Turns out I had 2 separate small issues affecting determinism on fixed seeds:

• An iteration on sets of location enum items
• An inline cast of a set to a list in get_filler_item_names

In both cases, I made sure the operations happened on lists instead to make sure the order is deterministic.

Before

2 generations with fixed seed 12345 would yield completely different item placement, progression items included

After

2 generations with fixed seed 12345 yield identical item placement

World tests are still green after the changes.

[ScipioWright]

Very simple change of sets to lists. Still weird that the itempool gets sorted by core, leading to this being an actual issue in the first place.