The law of diminishing returns applies here, as the efficiency rises, so each gain becomes ever smaller proportionally, so there's very little more at present that could make a dramatic difference electrically. Improvements in the metalurgy could bring us fractionally more efficient windings, but it would need a breakthrough in superconductive materials to make a big difference here.So...if cost were not an option, how close are the current hub drives to being the best current technology can come up with? In other words, are they already 80% there, 90, or more in your best guess?
Small mechanical gains in transmitting the power through to the wheel may still be found, but overall there's nothing on the horizon that would permit 90% to be reached, although marketing and advertising departments reach and surpass it without any problems.
Designers will get more clever in the way the available power is used in a bike though, so the overall package will show apparent advances to the user, even though they will be perceived rather than real. That doesn't matter of course, as long as the customer feels a gain, it's real. This is already happening.
.
Last edited: