They feel odd at first, but you adapt quickly. Even with minimal adaptation time, nobody has been able to measure a significant difference in any performance metric by changing crank length by quite large amounts. So, people don't have an optimum, as far as anybody can tell, and if they do nobody knows what it is with a precision of ±1.25mm, which is what the offering of cranks in 2.5mm increments would suggest. Of course, this much should be pretty obvious when you look at the huge range of people who are perfectly happy with 170mm cranks; if there were some ideal crank length, you would expect it to vary more or less in proportion to some morphological measure, yet people from Emma Pooley (5'2") to Theo Bos (6'3") successfully use essentially the same cranks.
Would smaller cranks not be more aero and also provide less of a dead spot due to the smaller distance they have to cover?
Would smaller cranks not be more aero and also provide less of a dead spot due to the smaller distance they have to cover?