Okay, let's see what kind of logical reasoning we can work into this situation.
People have been arguing that I should Guess 0, because if the answer is "yes", then there are 4 confirmed IC's. But seriously, what are the chances of that actually happening? This is RMM, and there are 1 or more likely 2 scum in the remaining 7 players of Reverse Timeline. Add to that the fact that there could potentially be 2 scum factions (which would probably mean 2 or 3 out of 7 are scum) or there could be 3rd party players involved (again, I'd guess probably 2 or 3 out of 7 are non-town). I'm going to mostly assume 2 non-town out of 7 for the remainder of this post.
As I said earlier, we need to sort of agree on how many scum (or non-town, actually) are most likely on my wagon (and we need the 2 most likely numbers). If there are 2 total out of 7, I'd put the chances of 0 on wagon at less than 10%. If a wagon is going to build quick, as it did, it's very possible scum would jump on it at some point, not wanting to be the last one to hammer a town, so they would have gotten on earlier. And especially not expecting it to go to hammer so early, it's a free place to put a vote. Mathematically, of 7 players and 5 town, there are 5 ways to pick 4 town and 35 ways to pick 4 random... 5/35=14%, but again, with the opportunity there, I think the chances of 4 town being there is much less. (Also consider if there are 3 non-town total, chances of 4 town on my wagon is 1/35=3% chance!)
I think the chances of 1 non-town being on my wagon are 30-50%. Again, assuming 2 scum who know each other, maybe they did go one on, one off, or maybe they didn't. We don't really know. We also didn't leave it long enough that they could watch, plan, coordinate (whether or not they have day chat), etc.
I also think the chances of 2 non-town on my wagon are about 30-50%. I think 1 is mathematically more likely, but to find out there are 2 would not be unexpected. Not expecting hammer to lock in their votes so soon, it's quite possible that 2 scum put their votes on, intending to take them off later if needed. (Add to that that if there are 3+ non-town in the 7 remaining, the chances of 2 on my wagon goes up significantly, whether or not they know each other).
And finally, the chances of 3 or 4 non-town on my wagon are low, in my opinion - maybe around 10% again (though possibly less). I don't see a reason to believe there is an overabundance of non-towns (it's not BM, and I know I'm town, so there are definitely town in the game). Possible, yes. Likely, no. And as roles become known through the game, if there are a lot of 3rd party or you find out that there are 2 scum factions, then you could go back to the results of this and consider if maybe there were 3 on wagon.
If I Guess 0, the potential reward is great (4 IC's), but I give it 90+% chance that the answer is "no". And what does that tell us? That there's at least 1 on my wagon. That's useless. If there's 1, it's a shot in the dark (with role powers and with exiles) whether you look on or off my wagon next Reverse day/night - it will fully depend on how many are still alive at that time and what else is learned (if no one dies, 1 in 3 off wagon, 1 in 4 on wagon, again assuming 2 total, probably means looking off wagon). But if there's 2, then you definitely want to look on wagon.
If I Guess 1, an answer of "yes" is not very helpful (as described above) - non-town is likely split 1-1 on and off wagon (depending on how other parts of the game go, I'd probably look off-wagon first, since there would likely be at least 1 in those 3). But if the answer is "no", then you know there's either 0 or 2+ on wagon. Maybe right now that's not helpful in some of your opinions. But I'd strongly wager on there being 2+ in that case (and if as the game progresses, there does seem to be an indication that they could all be town, rather than 2+ non-town, then that polarizing distinction can be helpful then too!)
If I Guess 2, an answer of "yes" can be fairly helpful. If one is scum and the other is 3rd party, it might be a bit harder to locate. But if they're both scum, it would make it easier to pair them off. But even if the answer is "no", that makes it fairly certain there is no more than 1 non-town on my wagon. You can look elsewhere and leave that last 1 (if there even is 1, rather than 0) for later. Out of 4 people on wagon, knowing that either 0 or 1 is non-town pretty much tells me you don't need to look here until LYLO or similar. And if for some reason, there are 3+ non-town on my wagon, then again, it should become fairly obvious as the game progresses.
Summary:
If you want a 90% chance that my result tells us pretty much nothing useful ("there is at least 1 non-town in the 4 people on my wagon"), on the tiny chance that you get 4 conf-town, then I think you're pretty scummy as you want me to most likely waste my role.
If you're okay with either confirming 1 on my wagon, or using your reads and the progression of the game to distinguish between 0 or 2+ on my wagon, then picking 1 is the right choice.
If you can comfortably lump (0 or 1) non-town on my wagon, and leave the 4 of them mostly alone until near the end of the game if the answer is "no", but getting a "yes" to 2 would be super helpful, then picking 2 is the right choice.
I don't yet know what I want to pick. I want the best odds for town in the game as a whole (that can be partly useful upon reveal, but potentially more useful as the game progresses). We can't guarantee to get a "yes", no matter what we pick. So how can we maximize the usefulness of a potential "no"?