I forget my mech-e skills but those large oscillating masses (the up down linear motion) would seem to be hinderance to efficiency. In a bike almost everything

I forget my mech-e skills but those large oscillating masses (the up down linear motion) would seem to be hinderance to efficiency. In a bike almost everything is moving in a circle on a bearing, it's the legs that are inefficient. Another thing is he's not really accounting for lifting the pedals like you do with clip-ins nor back forth motion you do when you get better at biking. I wonder what the torque curve of a trained biker looks like. I'm assuming a bunch of folks here have that in their training data set. I'm just here reading HN in the shade outside relaxing after walking the course yesterday. :)