For example, if I were to sign up and bet against Yang at 5% odds, even if I won my bet, I'd make $0 profit because my entire withdrawal amount is taxed at 5%. If Yang's odds were set at less than 5%, I'd actually lose money even though I won my bet!
Given this, the market for low likelihood outcomes is extremely inefficient and it's very plausible that these odds are being gamed by candidates vying for attention.
The weaker form of your argument "Without the caps it's biased toward people who can make big bets" is the entire point of a betting market.
Edit: The one edge case where it might happen is precisely this article. Buying your way to say 5% after some organic positive press might be an effective way to buy buzz at relatively low cost. It's a small one time boost for a campaign though.
But I get the feeling nobody really understands campaigns well enough (especially these long, grueling ones) to justify the time spent on these small-calibre issues.
Trump: +175
Biden: +650
Beto: +900
Yang: +2200
Buttigieg: +2800
Warren: +3000
Klobuchar/Booker: +3300
Hillary: +10000
The Rock: +12500
Jeb: +15000
(payout for a $100 bet)
Also interesting to mention the crossover between Yang supporters and the crypto community: https://breakermag.com/five-reasons-crypto-fans-are-joining-...
Ambiguous markets are ... not good.
Without the freedom for real amounts of money to come in and exploit mispriced bets, it's practically a straw poll of PredictIt users. The PredictIt user base has some obvious demographic biases that warp the odds.