I believe that the increase in limits for Chinook was because the catch was actually slower than projected. The initial limits were set to get the fishery to Labor Day. Since fish were not being caught as fast a modeled they upped the limit.

Interesting concept. Use the best data and models available to craft a fishery. A run of X should give a catch of Y. Then, when the catch comes in slower we actually INCREASE the limit? If the models are so good, doesn't catching fewer fish mean the run is smaller? Leastways, that's how it works in net fisheries. Apparently in sport fisheries, the fewer fish you catch the bigger the run.