I agree it is a challenging vision to contemplate, WMY - but I think the intent of hatcheries should be to work themselves out of a job - even if that doesn't currently look possible. The goal should be to return the wild stocks to some semblance of natural self-sustaining levels - and we can certainly argue about what that might look like - but I believe that should be the stated goal.
Along with a goal should also come benchmarks - or what might be called "measures of success". I really don't think that the hatch-out rate comes anywhere's near being a benchmark. I think hatcheries should be fully supported with enough capacity to be able to develop and assess what a "measure of success" looks like and use that as a goal - rather than just pumping-out numbers and hoping for the best.
Admittedly, there are unfortunately few models of that kind - or intensity of planning and data. The Cowichan might be the only exception to that rule I can think of or know:
http://www.pacfish.ca/Cowichan/
and more on hatchery impacts:
https://s3-us-west-2.amazonaws.com/...n/Hatchery-Impact-on-Wild-Salmonids+Bakke.pdf