So with regards to some recently submitted IL runs, it was discovered that running the game on the Xbox One results in a lower IGT than that of the video recording, and by extension of this, a lower IGT time compared to RTA. From the runs submitted, it can be extrapolated that the timer is slowing down to about 99.78% - 99.85% of what it should be (capture method doesn't appear to make a difference), as extrapolated from runs 100-120 seconds long. In practice, this is estimated to be resulting in a timer save of about 0.15s per minute.

2 possible reasons for this. One is that the Xbone is actually running the game ever so slightly slower than intended, meaning cars would accelerate and turn slower at a slightly lower rate. However this would give a theoretical advantage in ILs of having better reaction times by playing in a slowed-down version of the game, but I'd argue that the slowdown effect would be so small that it wouldn't be humanly perceivable and thus not at all helpful. The other is that the game is running at the same speed as PC, but the in-game timer by itself is actually running slower, which would actually quite a significant advantage for IL runs as it would be a free 0.15s per minute for absolutely nothing. In that case Xbone runs would need to be retimed to RTA for fairness.

Unfortunately it seems very difficult to test which of these scenarios is actually happening, due to the non-deterministic nature of the physics engine preventing the creation of two exact runs to be created on the different systems, as well as not being able to test replay files between Xbox and PC. The former scenario of the game overall slightly slowing down is most likely, given that this is an older generation game running on newer hardware and that tends to be a thing as seen with other games, however of the evidence I have reviewed it is entirely possible that the timer is just running slower but the acceleration and turning rates are the same, however the evidence is nowhere near conclusive enough.

Regarding ILs, for the time being I'm going to approve runs under the assumption that the entire game is running slightly slower, as this seems most fair to those running on those systems as well as being a lot easier to verify. However it is worth keeping in mind that if the situation is actually just the timer misreporting then these runs will need to be retimed to RTA.

For full speedruns, scenario 1 would mean that there's a slight disadvantage to running on Xbox One over PC (combined with the other disadvantages) that should be noted. Scenario 2 has no real effect to these categories. So overall no real issue here.