That is some interesting insight riverguy, and something that does not seem to have been addressed anywhere. So, if the department suddenly stops planting smolt, like they did on the Snoqualmie system, without first reducing the predator rate, the natural balance is offset. Predators that feed on 250,000 hatchery smolt suddenly see their food train gone. That first year everything left gets hammered. What might make it worse is when a system has staggering migration rates and pumps up several species and leaves the other untouched. Does anyone have a time table for the migration rates of different species? While it may seem odd, perhaps releasing steelhead at the exact same time as another species may encourage them to flush out quicker and face less individual pressure. Perhaps the lower stream quality, along with smaller plants of steelhead, leads to an outward migration all at once that allows a larger predator pool to take advantage of a feed source when it naturally low. An artificially high predator pool being fed by hatchery chinook and Coho and large pink and chum runs, surviving the rest of the year by wiping out whatever is left available.
There are a ton of different outcomes that come to mind when I think of the implications of an out of balance ecosystem. Something that should be considered when looking at release dates and outward migration. I can easily see how a complete cutoff of a hatchery run could instantly doom a wild run.

Is there anyway they could try doing the several of these releases next year and see how much difference timing and smolt size matters. Also, perhaps taking a few and transporting them out past the sound. I know it takes money but perhaps it can be raised based off the current results.