Sometimes you just have to try an experimental piece for a new writing technique or some such. Even if you're chiefly interested in writing a story -- Scott McCloud put his thumb on it when he pointed out that great story-tellers have often been technical innovators as they honed their tools to tell their story effectively -- and even if it's nothing new in literature, that doesn't mean that you can master it without experiment.
But then you come to the point where you have to look back. Did this experiment work?
If not, can its contents be salvaged from the ruins and rewritten in a less experimental form?