Thanks!

Hmm, yeah, the swiping UX seems to be pretty tricky to get right. Initially I didn't have the percentage but some early testers highly recommended I add it for two reasons:

1) They couldn't see well and found the addition of the percentage was helpful in clarifying their swipe value.

2) It made it more obvious that the magnitude of swiping mattered, as opposed to swiping just being a binary yes/no option.

The two concepts idea is interesting. I'll have to think more on that. I think it might be tricky to implement though, because not every topic has a natural converse.