I agree with most of what you wrote, and I suspect we might agree about most things, and to me there are some very important aspects of the context that are missing, and are needed to give any sort of coherent overview.
In this sense I agree completely with your sentence “Obviously, to understand humans you need to understand our evolutionary past”.
Three major concepts that need to be explicitly added to the mix:
Agree completely that context is critical, and one also needs to take a deep time view of context, and the sorts of things that have occurred over deep time – some high probability low impact, and others low probability high impact.
When you look at the sort of things that can happen, very rarely, like super-volcano explosion followed by years of growing season failure and mass starvation, then the idea that people “accumulate resources far beyond those required for survival” isn’t as sensible as it might seem to many.
Those tendencies to hoarding can be very costly, and they can come with very high payoffs.
You mentioned our liking for sweet things, and that is perhaps the worst heuristic hack present in today’s context – where sugar consumption is the single largest risk fact in diabetes, heart disease and cancer.
You mentioned context many times, but no explicit reference to the major modalities of competitive versus cooperative behaviours. In times of sufficient abundance for all, we come pre-primed with a large suite of behaviours that strongly support cooperation between individuals. And in contexts where there is insufficient for all to survive, we all also carry a suite of highly competitive behaviours.
Humans are both, and which gets expressed is very much a matter of context.
Perversely, as people do get sufficient resources to not need to compete, often the tendency to hoarding can take over and become the dominant strategy present if the more cooperative social strategies are not strongly reinforced.
So it is a many levelled and many dimensioned set of contexts and probabilities to action. Any given context may result in hundreds of different heuristics from our past (genetic and cultural) contributing to the probabilities that lead to whatever action we actually take. And there can be many levels of context present simultaneously, and we can learn to consciously shift weightings between levels or between competing contextual interpretations within any level.
So yes our deep evolutionary past is important (genetically and mimetically) and our conscious choices and our application of intentional will to the control of our own internal tendencies to both identify and classify contexts, and to modulate the responses appropriate to those contexts, can have a huge impact on who we get to be in reality.
Evolution matters, and we each have the individual ability to evolve who we are beyond the default dictates of our genetic and cultural past, and we can only do that by becoming conscious of as many levels of those defaults as possible, and retraining our neural networks accordingly.