The Forecast Eval tab in the forecast model says you might be right. We make escapement ( some streams ) when the run forecast is below actual returns. If you look at the Eval graph over the years it sorta cycles up and down but it is an window of around 4 years each way more or less. In other words in years of an up cycle the forecast lags behind actual return but in years of down cycle declining run size forecast over predict returns. The Eval tab only goes to 2012 but I imagine the pattern holds true.

The lowest wild Coho return I could find was 1991 with the Chehalis at only 6615 / Hump 1005 which gets you 7620. That years return was in a 10 year window with a high of 44k to the low of 7620. The returning adults off the 1991 brood was 17k and change. So the simple fact while Chehalis Coho our prone to a pattern it does have years of rather large or small returns that are not anticipated.

By not allowing the tidewater tribs to make escapement when we have an up cycle we do not benefit because of the reduced brood in those tribs. Make escapement consistently and up cycles are years of abundance vs what we have now is hanging in there good years or bad. Like the man said " Houston we have a problem "!