I would only argue with the 1:3 odds, which I'm not sure how you ended up with.
The worst case scenario would be a game with an unpredictable (50-50%) outcome - and that would still make it better than any dice sites with a house edge.
Odds =/= chance to win
Let's take the next game, "cats average follower count is higher than number of tweeters".
There is .005 bet on "true" and .005 bet on "false". Let's assume that there will be no more bets if I make a bet (a fair assumption).
If I bet .005 on "true", I am risking .005 to win .0025. And then of course the 1% gets taken out of everything. So I am really risking .005 to win .002425.
I, of course, would have those same odds if I bet on "false".
So on a pure 50-50 shot, half the time I win .002425 and half the time I lose .005. That is certainly not better than any dice site.
Of course the equity drops if I want to bet more than the bare minimum, because of the nature of parimutuel betting. This is where the 1:3 comes from, as I have no interest in betting .005. Betting any sort of decent amount would give odds of 1:3 or worse, because of the low volume and the fact that you've seeded the pools on both sides.
I'm thinking I will use part of my budget to add more bot votes. I suspect most visitors are like you - it's not worth their time to play just to win small amounts (and you would definitely win considering most bets are posted randomly, by bots).
It would be more attractive if you stopped the bot voting altogether and just seeded the prize pools. When you place bets on your own game, you are reducing the amount the players can win by ensuring that you always get a cut of the pool. Instead of betting .1 on one side and .1 on the other, just add .1 to the prize pool (like a pick 6 carryover, or the cherry bonus on bettingblocks)