I prefer walking on steep hills.
a while back Pete & I were measureing a steep hill section of a course. Pete got off and walked the bike holding down on the handlebars, loading the front wheel.
I rode up the hill and although I thought I got the better measurement, Pete beat me on that mile segment by about 3 meters.
I re-learned the Old Bull/Young Bull lesson that day.
An alternative would be to ride down the hills but that makes for extra work and can be difficult to convey to a certifier. Does riding DOWN hills not have some effect on the measurement due to added speed, riding the brake, or not pedaling?
I suppose you could calibrate walking the bike, then note sections where the bike was walked. This would probably be best if the course was mostly hills.
About a year ago I added a 70cc gas engine to the bike I measure with so hills and long hot days are less of a problem.