I'd argue the most interesting part of this piece isn't what it says about Paul Graham, but rather the observation it makes about writing. I think about "It Turns Out" all the time, and it's virtually never because I'm in that moment caring about something Graham wrote.

The rebuttal is especially interesting because it simply let's the actual usages of the term speak for themselves. It turns out (ha!) that the Cambridge example is the only case that supports the OPs case.