Which edge cases deserve a test in a small project

The ones that already bit you, and the boundaries: empty, one, many, malformed, and the largest input you claim to support. That list catches most regressions without turning the suite into a second implementation.

At ten times the size the bottleneck moves from the operation itself to what it competes with: memory, locks and the people who have to run it. The approach stays, the batching gets smaller and the schedule matters more.