There are some great questions here! I'll try to answer them with the caveats that 1. I'm
not an expert 2. I'm not a great science educator 3. given some of the posts in this thread I'm not sure this is a receptive audience. I'll do my best to provide links where I can so you can get more information presented better than I'm able to. There are
a lot of factors that are at play here and a deep dive (ahem) into all of them would be "publishing your Ph.D" level writing, not a "sportfishingbc" post.
I'll attempt to paraphrase part of your question to "What clues does science have to explain why fish eaters don't turn to seals when chinook abundance is down?"
In environmental science/ecology, there are "generalists" and there are "specialists". Generalists are easily adaptable to changing environments while specialists are finely tuned to specific environments and don't handle change well. Further reading on generalists vs. specialists:
NG link Albert.io link Fuse School video
A local example of a generalist would be a glaucous winged gull. These birds are able to thrive everywhere from downtown Vancouver where they'll eat french fries and fentanyl needles to the most pristine areas of the BC coast where they'll eat herring, crabs, and other natural prey items. Put a strip mall on Cape Scott and the gulls that live there will switch from herring to french fries in a day.
A marbled murrelet would be an example of a specialist; they require specific feeding and breeding areas to survive. Log the old growth that they need to breed and they're unable to move to other habitats.
In
undisturbed environments where resources are not limited, critters can afford to be specialists without too much risk. Advantages to specialization include being able to exploit specific niches that others can't, therefore removing competition.
Looking back to some of the KW reference material that I've already linked to, you will see that genetic studies show that Bigg's and "resident" (fish eaters) are separated by ~350,000 - 700,000 years of evolution, and up until recently (before west coast colonization/industrialization) there was no problem to be specialized as a fish eater or a mammal eater because there was plenty of both to go around and neither population was directly competing with each other. EDIT: and back in the day when chinook were abundant and huge (~100 pounders in the Elwha?), the cost-per-chase differential wasn't that big of a deal.
Salmon used to be bigger,
but not so much anymore,
which is a bummer.
Once disturbance comes into play the specialists will begin to suffer. So, once the ecosystem around here started to see the effects of industrialization (clear cutting, dams, overfishing, urbanization, habitat loss, etc. all of which took place in the last ~150 years or so) the specialists can't adapt fast enough. Keep in mind killer whales have about the same lifespan as humans, so ~150 years is only a few generations. That's not long to undo half a million years (+-) of evolution.
"One of the key features we assign as a marker of intelligence in animal species is ability to adapt to varying conditions and availability of food. Are we off the mark with orcas, or is there something else going on?"
There's zero question that killer whales are highly intelligent.
Don't take my word for it,
there are many examples of how intelligent they are. The cultural differences are especially important to take into account. Captive mammal eaters have learned to eat fish (or die), and captive animals from different populations have learned each other's dialects and can mate and produce viable offspring. But in the wild those sorts of shifts don't happen. You're born into your ecotype and you'll die your ecotype.
You mention "The mammal-eating orcas have it figured out" but they're just the lucky specialists in this case, and it wasn't necessarily always this way. When KW studies began in the 70s, there were only a few dozen known mammal-eaters on the BC coast, while there were approximately 300 in the SRKW and NRKW populations. Keep in mind pinnipeds were subjected to serious culling efforts in BC and elsewhere, with the harbour seal population in the early 1970s being estimated at fewer than 10,000 animals and Steller sea lions fewer than 4000 animals. That's not a lot to sustain a population of Bigg's. But once culling stopped and the harbour seal and SSL population started to rebound, it's not surprising that Bigg's population followed. The NRKW population, which generally lives in more remote, less disturbed areas of WA, BC, and SE AK, also seems to be doing much better than the more urban SRKW. We don't have the luxury of knowing how many Bigg's or resident types there were before human disturbance but it's likely that there were more of both types.
"Being fixated on a specific food source even when other sources are available seems counter to evolutionary survival. What am I missing here?"
What I think you're missing here is that you're applying a human timescale, where we've radically altered the environment in an extremely short amount of time, to an evolutionary process that's taken our local killer whale populations hundreds of thousands of years to get to. They can't evolutionally turn on a dime.
Long story long (again) I'll try to summarize here:
"What clues does science have to explain why fish eaters don't turn to seals when chinook abundance is down?"
Our local killer whale populations evolved to be specialists and can't just switch food sources on a whim. Their specialization is literally baked into their DNA.
"One of the key features we assign as a marker of intelligence in animal species is ability to adapt to varying conditions and availability of food. Are we off the mark with orcas, or is there something else going on?"
We're not off the mark but we have to try to understand and apply killer whale intelligence in the context of their evolution and environment, not our expectation.
"Being fixated on a specific food source even when other sources are available seems counter to evolutionary survival. What am I missing here?"
Human disturbance timescale vs. evolutionary timescale.
Hopefully this helps!