Assuming the current probabilities of those markets are accurate and Scott probably only reviews the three or four with the highest chance (and the others therefore resolve N/A), 35% seems about right.

There's a not-quite-arbitrage opportunity where if the top 3-ish markets here, multiply their probabilities, then add the probability of this one - if it's greater than 100% you can buy no in all 4 markets and make money with very high probability

If not, then maybe it seems like Scott's setting the bar too low. Markets indicated by this post:
https://manifold.markets/ScottAlexander/if-i-review-rick-perlsteins-nixonla
https://manifold.markets/ScottAlexander/if-i-review-joseph-stiglitzs-whithe
https://manifold.markets/ScottAlexander/if-i-review-mark-brinells-the-searc
https://manifold.markets/ScottAlexander/if-i-review-edward-teachs-penelopes
https://manifold.markets/ScottAlexander/if-i-review-one-of-rene-girards-boo
https://manifold.markets/ScottAlexander/if-i-review-vaclav-havels-the-power
https://manifold.markets/ScottAlexander/if-i-review-bruce-finks-a-clinical