I think that's right for just picking a different PRNG seed. When you start looking at modifying the search heuristics, mutation operators, or other parts of the "strategy", you definitely start finding different bugs, though.
Yes, absolutely. Different strategies can lead to vastly different results, often it's subtle things.
Good example: There was a bug I found in openssl that Libfuzzer was unable to find. The Libfuzzer developer was quite interested in this and has now adopted new mutation strategies:
https://github.com/google/sanitizers/issues/710
Good example: There was a bug I found in openssl that Libfuzzer was unable to find. The Libfuzzer developer was quite interested in this and has now adopted new mutation strategies: https://github.com/google/sanitizers/issues/710