It hasn't been tested yet

Info is a little uncertain right now.
The only test of FDS against SDS (and emu) has been on PAL, on a 30k, and anyone who knows about PS2s in speedrunning will know that 30k is a terrible representation of FDS.
But with 30k as FDS, the results are (after ~2 minutes of in-game time):
SDS, FDS, and ePSXe 2.0 have the same IGT to the second
ePSXe 1.9 loses IGT time
ePSXe 2.0 had the fastest storybook time (due to little lag), although FDS had the faster RTA time if you count the black screen into the loading screen
Similar to those differences, ePSXe 1.9 was faster than SDS for storybook, but SDS was faster when adding the black screen.

Also, one other difference I found when testing was that PAL is faster IGT than both NTSC-U and C, and NTSC-C might be slightly slower than NTSC-U.

Here are the results as a table if that's a bit confusing

I'm not sure if we really want to go ahead with this quite yet, since, like I said, 30k FDS is bad - much rather a 70k test but my 70k doesn't want to play nice. But for now, ePSXe 2.0 is allowed (and 1.9 but since that's slower, why use it?), but PAL runs using IGT might be separated (although it doesn't seem right to do that, since there's no gameplay difference and what's fastest is fastest).

Why was PS2 banned on the first place?