For instance, imagine an exceptionally intelligent and hard-working individual who leaves a high-paying job in finance or tech to work on a startup. Have they suddenly become unproductive because they have switched to a role that pays less (in the short term)?

This is a very uncharitable interpretation of what I said. Their productivity in the startup has to be based on an expected value calculation. If the startup accomplishes nothing and goes under, then I think it is fair to say that all their hard work was unproductive and they would have been more productive selling their labor to someone who knew how to do something useful with it. If their startup makes $Billions then all the work they did to bring it to fruition was very productive. Ahead of time you have to guess at the distribution of possible outcomes and base expected productivity on that.

I said that if lukeprog's current activities somehow yield big dividends in the future then I will have been wrong about his current productivity. I just don't think that is very likely.