I suspect this is also a "HIVE" mind at work too,not the actions of a single bee
I think the problem is of individual bees, so if it is collectively solved, the result is individually required so no economy there. The observation IIRC, is that bees, who stick to a single flower type on a given day, take a maximally efficient path when going flower-to-flower gathering pollen.
Should we follow nature in our designs? On a physical level, yes and no. Animals have typically rotary DoFs in their limbs and fingers. If we are making a fluidically driven system, linear extensors are as easy as rotary ones. No good reason not to use a piston just because it isn't 'natural.' A peristaltic pump is quite 'natural,' but only situationally appropriate.
On the other hand, when I read about and meditate upon hormonal models of behavior, given the evolutionary context, often I have trouble imagining alternative structural possibilities.
I find modeling Turing's morphological models to be quite rewarding. Also, I am a big fan of Lindenmeyer grammars.