I just ran my stats across all professional games recorded by gokosalvager.
Here's the google doc:
https://docs.google.com/spreadsheets/d/1fEm6TtcqLWrqq5ksaOdEXnXEn9FfDWaShpwtnuxLvWA/edit?usp=sharingNote that games present is games where its in the supply; the "discrepancy" with Gear isn't a discrepancy: I lost a game where I bought it out of the Black Market deck; but haven't lost with it otherwise.
By looking at the base cards, I have a score % ((1 * wins + 0.5 * draws) / games played)) of 65%. The non-Adventures card I have the highest score percentage with is Great Hall, at 85%. Weird. I tend to buy it in those games as well.
As for actually buying the cards, I've apparently won every game (extremely small sample size) in which I've bought Stash, Harvest, Coppersmith, Duchess, and Bureaucrat (again excluding Adventures, my win rates are very high with those cards). For cards that I actually have played > 10 games with, buying Embargo, Herbalist, and Scrying Pool all lead to 90+ win rates for me in rated games. Interesting.
On the flip side, my win rates when buying Count, Native Village, Trader, Horse Traders, Advisor, and Workshop are all <40%.
By the way, my extremely simple methodology of computing this means that Knights aren't calculated correctly for being bought, but they are for being in the supply.
Also, no, I can't easily extend this to anyone who asks.