Pardon me if I politely decline to drink the fishery observer Kool-Aid. Fact is, I smashed the mug, into which someone had poured catch share Kool-Aid, in the fireplace long ago.
Observer programs are not inherently evil, but they're not inherently sensible, either. Billeting qualified scientists on fishing vessels is often impractical and never cheap, regardless of who is picking up the tab.

A NMFS observer checks the dimensions of a net and its catch. NMFS photoI don't quarrel with the collection of fishery-dependent data; it's just that observers are an expensive way to gather it. I realize that observers are scientists gaining valuable insights in the field. But much of the information is within the grasp of the average fisherman, so let the fishermen gather it at sea and the scientists deal with it ashore. To the extent that they collect biological data that would ordinarily be beyond the scope of a deckhand's duties we should think in terms of innovation and not resign ourselves to what a biologist's job has always been.

I am also skeptical of observers as compliance monitors. Call me naïve, but I am not inclined to view fishermen as lawbreakers or cheaters. Besides, we know where folks are fishing and with a modicum of shoreside enforcement we can be certain of what they're landing. That said, bycatch, particularly in some high-volume pelagic trawl fisheries, is an issue that needs to be addressed. Seasonal closures are one method of accomplishing this, but there are times when observers may represent another. At the scale at which the pelagic trawlers operate it may be easier to justify an observer's limited presence.

The answers to the challenge of fishery management will seldom be certain, but they need to make economic sense with respect to all resources, by which I mean the ocean's, the fisherman's and the taxpayer's.