At first I thought to judge this submission, but after a quick look I noticed a problem. I'm not going to claim this before I discuss and figure out how to judge it.
The thing is, the
current publication uses BizHawk (NESHawk core) while this submission uses FCEUX. Both emulators are allowed for making submissions, however in this case it's troublesome for me to compare the two movies, because FCEUX generates less lag frames and different RNG, making it impossible to say if this improvement is more optimized or actually benefitting of the emulation differences.
One could argue that we there is no certainty that BizHawk is more accurate than FCEUX, and thus both emulator should be considered equally valid
(unless someone provides a real console verification). However, the fact is, we already know of another FCEUX movie (thanks
mtvf1) that we can compare this new submission to, which can be downloaded from
here. If we disregard the starting cutscene, that FCEUX movie is about 30 frames faster than this submission.
By the way, using different difficulty modes shouldn't be an issue, since TASing for this game pushes the gameplay to the point that harder difficulties don't actually result in a more challenging TASing process, but affects gameplay just because of different enemy spawn positions, and thus a TASer may pick whatever difficulty mode results in a more favorable outcome for fastest completion. Note that this submission used easy mode, the published one used hard mode, and the old unsubmitted movie used hard mode 2nd quest (still no idea what's different from hard, apart from the lack of intro cutscene).