Haven't ridden a Scram, so don't know the practical difference. But I agree, it's supposed to favor a quicker rise in the torque curve (low-end torque) but at price of a quicker drop-off (thus less high-end power).
As to the geometry: Imagine the piston is the second hand on a clock. One full cycle of suck-squeeze-bang-blow is two revolutions around the clock. If we start with the second hand on the 12, then the next half-revolution (180 degrees) is suck, followed by a half-revolution of squeeze, with the spark coming roughly at the 12 on the next time around. The next half-revolution another 180 degrees) is bang, followed by the final half-revolution of blow and then we're back where we started. Clear enough?
Now, on the 360 degree twins, the two second hands are always pointing at the same place on the clock face, but they are one full revolution (360 degrees) out of phase, so that when one is starting to suck, the other is starting bang, and so on. So one is doing suck-squeeze-bang-blow and the other is doing bang-blow-suck-squeeze.
In the 270 degree motors, when one second hand (piston) is pointing at the 12, the other is pointing at the 9. So when the one at the starts to suck, the other is halfway through squeeze. Or, to talk about the actual firing, when one of them starts the bang, the other one is halfway through blow.
If you break each of the 4 strokes into 2 parts, so that there are a total of 8 parts per full cycle, then, representing the left cylinder with an L and the right one with an R, and no cylinder with a -, then the firing pattern for the 360's is like this:
L---R---L---R---L---R---L---R--- etc.
whereas for the 270's it's like this:
L--R----L--R----L--R----L--R---- etc.
Clear as mud! I have no idea why this change has an impact on torque curves.