1. I'd particularly expect many people to have good tacit rationality without having good explicit rationality in domains where success is strongly determined by "people skills." This is the kind of thing I expect LWers to be particularly bad at (being neurotypical helps immensely here) and is not the kind of thing that most people can explain how they do (I think it takes place almost entirely in System 1).

2. When evaluating the relationship between success and rationality it seems worth keeping in mind survivorship bias. For example, a small number of people can be wildly successful in finance through sheer luck due to the large number of people in finance and the randomness of finance. Those people don't necessarily have any rationality, explicit or otherwise, but you're more likely to have heard of them than a random person in finance. But I don't know enough about Oprah to say anything about how much of her being promoted to our collective attention constitutes survivorship bias and how much is genuine evidence of her competence.

3. One setting where explicit rationality seems instrumentally more useful than tight feedback loops is in determining which tight feedback loops to expose yourself to, e.g. determining whether you should switch from one domain to a very different domain, and if so, which different domain you should switch to. IIRC there are various instances of well-known and respected scientists doing good work in one field and then going on to spout nonsense in another, and this seems like the kind of thing that explicit rationality could help prevent.

Agree that I am spending too much time reading LessWrong, though. I've been quantifying this using RescueTime and the numbers aren't pretty.