Even this thread has people saying that all that matters is getting something to market, which I think is indicative of one of the major problems of modern software development.
I mean the reason we write code in business is to make money and continue to provide features/fixes at an efficient rate to continue to make money. If you can do it right the first time within the confines of the time allotted, go for it. But if it comes at the cost of making money/spending time, we aren't doing it for an esoteric reason.
> continue to provide features/fixes at an efficient rate
The crappier a codebase is the harder this becomes