That's just about over. The state of the art in most areas of coding is getting so advanced that ad-hoc approaches simply don't cut the mustard. The competition is using an approach that is provably optimal, or provably within X% of optimal, or provably optimal in a time-quality tradeoff sense.
And there are so many strong solutions to problems out there now that cutting your own is more and more just foolish and hopeless.
I wouldn't say that the end of the ad-hoc programmer has come just yet, but it is coming. And that's depressing.