Perhaps a further fine tuning to more specific content can be gained by replacing ‘instinct’ with automatic behavior.
In that way some interveaning ganglia may morph into the more general cognitive function, .
That is what I get from Your pre-existent cognitive mapping of instinct.
Automatic seems more of a transitional function through learning rather then pure instinctive behavior, that may be viewed as some given at certain levels of morphology, leaving the learning process more like a series of validations of hypothetical summations at each level.
Which may beg the question of learning being a repetitive review of various newly formed coalesced or reified structurally reduced toward lower symbolic levels.
Not that lab demonstrations are not useful, but they only affirm the learned reduction through memory storage.
I would think a lot of lab work with the cognitive basis of dogs’ faculties are done to validate the hypothetical assumptions underlying them.
A lot of Russian research with dogs, and of course Pavlov setting the trend.
My approach bases is mode intuitive, and again begs where general hypotheticals about brain-mind function actually come from ,