Dispatch Optimization - CSP Power Tower

Dear Paul,
I have been reading about the SAM dispatch optimization algorithm, but this sentence that is in the official information is not clear to me:

"When you run a simulation without enabling optimization, SAM attempts to operate the system so that the power cycle output matches the dispatch schedule you define. If the turbine dispatch fraction for a given time step is greater than the minimum turbine fraction and the storage system is full, it will dispatch energy to the power cycle up to the turbine output fraction for that time step, and defocus the field if the field thermal power output exceeds the amount required."

This question was born from an old post of mine, where I showed that the turbine fraction did not adapt to what was specified because the storage was already full ( sam.nrel.gov/forum/forum-general/3519-results-in-tower-csp.html?start=0 ).

According to this paragraph, the simulation does not need to be optimised to fit the defined matrix because when the tank is at full capacity, the system defocuses its mirrors. Perhaps I am misinterpreting these words, as English is not my mother tongue.
Or maybe this defocusing of mirrors occurs for the next time step and not in real time, that's why the behaviour I showed in my old post occurs (unlike the optimised system, which manages to anticipate these changes earlier).

As you can see, in the example in my old post, the dispatch fraction after the tank is full is higher than specified. This contradicts what the paragraph says, doesn't it?
I would appreciate your opinion, please. And again I thank you very much.

Francisco