So now that a few runs have come out with the Preview builds with significant variability between the IGT vs RTA, let me try to explain a bit better what's going on, and why I am hesitant to make the switch;
IGT is a recording of frames; the game runs at 35 frames per second, and if it lags, no frame is counted, so that's a true measure of how many frames elapsed; this is why it drifts from RTA.
RTA uses GameMaker's current_time, which should look at your actual CPU clock, and therefore does NOT account for lag; theoretically, it should line up with an external timing program.
That said; if your PC runs slower, you'll have an "easier" time setting an IGT record, but a harder time setting an RTA record. This is why it's tricky to make the switch, I either end up favoring people with faster / slower PCs.
Is this worth creating separate Leaderboards for IGT vs RTA?
Thoughts? Suggestions?

I'll let others decide that. @hahhah42hahhah42 said he did feel he had an advantage using IGT for his runs time.

I have a few ideas for this, but only suggestions (and a bit of speculation which you can point out if it is invalid or wrong),

If we do use RTA runs for times, IGT and RTA should be in separate categories and here is why I think that.
1. I have a new(ish) pc, coming in at around just over half a year old, and it is very well suited to run games that are much more demanding that WWW, so with that said it makes my RTA very close to my IGT. Which is, roughly 3 seconds while hahhah42's RTA has about an 6 second discrepancy.
2. RTA runs on a timer set within your computer, not set on the speed of the game. Therefore the slower the computer the slower the timer, making the faster computer have a (in this case) big advantage from the start.
3. Looking at our runs, (in which hahhah42 does have a significantly better time than me) I see that even as early as 1-1 (while collecting 1st and 2nd leaves), and slightly less optimal movement, my RTA is still ~0.3 seconds faster. As where the IGT is accounting for how fast the game is running, and therefore putting the person running the game better in the advantage.
4. My last real point I want to make is that the advantage given to the runners with slower computers is a good advantage. But, I would rather them have the advantage of playing the game at a slower pace and performing better than a runner with a faster computer getting a flat few second advantage over the other runners. While runs may be easier for hahhah42 if he is running the game slower, there is a way of working around it, while I have the advantage of just pulling 3 seconds out of my pocket and there isn't a work around for that. IGT measures how fast you are running the game, and RTA measures the time it takes in real time to finish the game, which is inconsistent across all PCs while IGT is the same.

While I am leaning towards IGT, and all being said, the better runner is leaning towards RTA. Either way, or rather either means of timing the game, I am accepting of and will respect the decision. I will continue to run the game, if I have the advantage or disadvantage, and please, inform me if any of the points I made are invalid as I am not the most amazing when it comes to all this wacky technology bs.

The only thing I should clarify in your post, @jakersisjakersis, is regarding point #2; that the RTA should not be any different between "fast" / "slow" PCs; it's the internal UNIX time-stamp; responsible for keeping your system clock accurate. Your PC would have to be in pretty rough shape for this not to match up.
So; RTA should be consistent across all PCs, but the discrepancy between RTA and IGT will vary depending on your PC; the better it is, the less of a difference there is between them. But it's never going to be perfect; I really need to run the desktop version of the game with a timer, because I anticipate most of the lag is coming from the not-very-fast HTML5 functions which (I imagine) the dedicated desktop version can chew through without lagging.

Quote

IGT is a recording of frames; the game runs at 35 frames per second

Honestly, that makes me less opposed to IGT, provided we start converting the ms properly. I was concerned that something else was going on.

Well, I could amend the algorithm that converts the total tally of frames into Minutes:Seconds.Milliseconds; I think the only thing that's inaccurate about it right now are the "milliseconds", which could only be interpolated from the frame-count, for example; 7 frames divided into 35 = .2 "milliseconds"
Using IGT will never equate to RTA, I can't think of a formula for consistently converting between the two because it depends too much on the PC running the game