I don't buy this argument for software in the least; it presumes this zero-sum tradeoff between "effort to build" and "quality of outcome" that I just don't bel

I don't buy this argument for software in the least; it presumes this zero-sum tradeoff between "effort to build" and "quality of outcome" that I just don't believe actually necessarily exists in a medium of infinite reproduceability like software. In my experience what's really happening here is the "hypercard" problem. Namely, there is a tendency, perhaps subconscious, for programmers (or perhaps meta-programmers) to NOT make "tools for building" too easy because they're scared of writing themselves out of a job or a market -- or they're trying to create or retain some exclusivity. I'm not necessarily saying this is always evil, but it's always a thing to consider and is likely happening all the time.