In lieu of the recent MK64 mishaps and just general benefit to the community, I want to discuss the proof requirements for World Records and/or top 5 (or higher?) times. I think that we should update these requirements as there's no reason not to, with the bare minimum being at WR level. Most of the WRs now have video, and the only ones that don't are WC 2lap by Goomba in which iirc his hard drive had issues, but with Dolphin ghost replay, it is possible to recreate it, he just hasn't done so yet. The other one is WS Flap by Goomba as well, not sure happened to this time... and Andrew's RR 3lap 2'54'239, which he never bothered submitting saying he didn't want to since he didn't have video of the time. Of course, the current times that don't have video should be grandfathered in, but the times that have those supposed videos should either wait until their videos are released, or there should at the very least be an explanation for the lack of video. Most if not ALL of the current top players have access to a capture card and/or tools in which to recreate their PRs. These tools are incredibly easy to use and are as following:

Fixed Replay Camera is a new AR Code that as the title implies, fixes the replay camera to work like normal instead of the old Michael Bay-esque camera angles. example:
https://youtu.be/xInjyK61VGo?t=102

Dolphin Ghost Recreation, a tool made by SwareJonge, which from simply importing your ghost to your computer, you can recreate all your saved ghosts. example:
https://www.youtube.com/watch?v=k7etvz_yAzI
This Dolphin Ghost recreation tool can be found here: https://github.com/SwareJonge/MKDD-Ghost-Info/releases
with instructions here: https://github.com/SwareJonge/MKDD-Ghost-Info/blob/master/README.md

The biggest exception to this would be flaps, as you can't normally save flap ghosts unless you beat the ghost that you were racing against. However, this can also be bypassed by utilizing a separate AR code, Time Trials Ghosts Are Always Savable. These AR Codes are all listed in this: https://pastebin.com/AH4eD1Df

Another cool thing about this is that with these tools, we can also recreate former WRs so we can have better documentation of the past (for sites like mkwrs.com or WR history progression videos) from players like Kouider, Andreas, Richard, etc., if they could somehow provide some of their ghosts, but that's talk for some other time lol

This update on the proof requirements should be required now that we have access to these new tools and there isn't really a good reason not to have stricter proof especially at the topmost level of the game. Please let me know what you all think about this.  

I know the games and communities are quite different but I'd like to mention something I came up with for OoT leaderboards that I think has been working really well.

What you really want is proof above a certain skill level, not really proof above a certain leaderboard rank. Rank can often be a really nice proxy for skill level, but it doesn't have to be. 50 people could all get within 5 milliseconds of a record if they all wanted to grind one course for a few months, and then rank cutoffs for proof kind of break down. To solve that, you could have the cutoff be some time multiplier on the current WR. In OoT we use 10% but that is kind of a lot for this game. Maybe 1%? You already track PRSR as a stat, so it would be pretty easy. If someone wants to submit a run at 99% PRSR or better (or whatever cutoff you want), it needs video proof.

Short version is that the number of people who achieve a skill level doesn't change how impressive that skill level is, and so it shouldn't factor in to when proof is required. Let the time of the run speak for itself so that the cutoff is always relative to the best time anyone has achieved.

Hey Danny, PR/SR could also be an alternative, my only issue with it is that it'd be an ever changing position with WR improvement. That'd also be an issue with top x rank as well, but it'd be tons easier to track x rank as you would simply have to look at the leaderboards for each rank. PR/SR also might not work well with tracks like WS 3lap or BP 7lap where 1:02.3 is wr, 1:02.9 is 2nd place and that is already at the lower levels of 99% lol We could always start at top 5 requiring proof, and later on as more players start to reach the top, we can push this limit further to top 10 or etc.

I want to see Kouidier's BP flap

At the moment I'm pretty far removed from knowing what's possible and good what's not. I last played before a-tech and lately there's been lots of medium-ranked players getting strong times on various tracks and I just assume they're copying twitch strats. Someone who is active in the game would be better to judge if/when times look dodgy or if/when players should submit proof for something. Basically if anyone points out to me that something looks suspicious then I can probably act, but otherwise I mostly just accept whatever people send in since I don't know (unless there is an obvious typo, which I can correct myself).