I use the predictive horizon model. It's not that there's a clear division between first order and second order thinking, it's that some people's ability to model consequences covers a narrow area, a few people can model successfully much further out, and a tiny minority can make astounding predictive leaps that reach far into the future.

It's a continuous variable, not a Boolean.

But it means most of the population literally cannot see consequences that seem obvious to someone with a wider reach.

Asking "and then what" won't change this, because if your modelling ability is poor you'll just get a longer list of wrong predictions.

It's a mix of native ability (IQ helps, but it's not a sole determinant), domain specific experience, and experience with synthetic thinking - the ability to filter and combine multiple trends, as opposed to purely analytical thinking which usually looks at one or two.