As I understand it the way we decide whether something is used for time attack is if there is, after using whatever trick, still competition potential. This is why 8"25 m2s on SA2B are banned completely, for example, while DD languishes in freestyle.
Now I have two points to address, a time stop and a infinite time-scalp technique.
Time stop: Time stops have always been allowed in time attacks (Cannon's Core, Sand Ocean, Launch Base 2). No issue here. I might try it out soon though, I've never actually seen what this glitch does.
Infinite time-scalp: Even if you can get up to 99 (or 127) minutes, there is obviously still competition potential, as you can see by looking at the chart. Therefore a straight ban is not the answer.
The only reason I would consider banning this technique is because of the issues with the DC timing method. Even then, that's a game storage limitation and shouldn't impact rankings (see Final Egg rings for example). Note that the Marble Garden 1 ring ruling doesn't apply here because at some stage you do see the full time of 127 minutes.
Additionally, we can't ban things on a subjective basis ("this isn't fun to TA") because put simply, TAing is somewhat intended to be a frustrating exercise. This also leads me to suggest that it probably shouldn't be put in freestyle either (I don't particularly like the freestyle concept, but that's another story for another day).
Anyway, tracking for HS times is left with five options: (maxes under each regime in brackets)
1. Anything goes (GC 99min, DC 127min)
2. DC times should be recognisable from screenshot (GC 99, DC <102)
3. Neither version should be at a disadvantage (99)
4. No time scalping, time freeze allowed (GC 15?, DC 5?)
5. Time scalping and time freeze disallowed/moved to Freestyle (5)
Per my above statements, 5 is out of the question as time freeze is allowed by precedent.
Anybody who would prefer the disallowing of time scalping would therefore be in support of 4, which means they can't use the version disadvantage argument because GC has a huge one under that rule.
Pros: No more using two hours just to get near the top of one chart
Cons: Large (read: can't be caught up) version advantage to GC, revamp of charts, relies on subjective analysis of methods
Method 3, which appears to be the version we work under now, seems to work okay in practice but has a major flaw, which Psyknux recently exploited. If we simply limit the time that can be submitted to 100 minutes, then a player can go over 100 minutes and wait until 99'59"99 and throw the final shot at the boss. The only way to counterbalance this would be to suggest a "you can't at any stage go over 100 minutes in the level" rule... and congratulations, you've just created the most unworkable rule in the history of TSC. Change is needed.
Pros: Current system, has worked for three years, no real version difference
Cons: Abusable on DC, unworkable rule
Method 2 is similar enough to method 3 but additionally recognises that the version difference exists, while also ensuring that the time can later be proven. However, it's just as unworkable as the previous rule, as well as giving a version difference and is just plain silly.
Pros: Similar to current method, allows DC's additional method and can still be proven later on
Cons: Unworkable rule, version difference, not particularly intuitive
My preferred method is method 1. The only issues with it are proof and the character-breaking version difference. Proof is easy enough to organise, but the version difference is another matter. I think it's silly to impose a limit on DC for the sake of the GC players though.
Pros: Consistent with TSC's "if the game lets you do it, it's allowed" philosophy
Cons: Proof issues, unbeatable version difference (lucky Gamma doesn't count towards the total!)
So yeah. Debate over.