What does that have to do with what I said?

I've built large FE systems with millions of lines of code pushing tens of billions of dolllars in revenue (I was lead FE dev on that one).

The idea that failing to do a small project (that should be clearly represented in the training data) means failing even harder at bigger projects doesn't suddenly gain merit because I have a lot of experience and have worked on big projects.

On the Internet I see all the time people who are building equivalents of bike sheds criticizing people who are building skyscrapers.

Your first sentence is telling me you were in more of a bike shedding projects than skyscraper one.