I was just coming here to post about this bug. I was finding that if I micromanaged the superchargers by forcing some to be used, I could get faster travel times than the original solution.
One other thing I noticed, I think your travel speeds come from Google's API? I've noticed that it is a bit low in many areas and this translates into much lower wh/mi consumption. For example, in west Texas on I10 where the speed limit is 85mph, ABRP indicates an average speed of 60mph. I realize this is probably just pass through info for ABRP and not necessarily a bug. Just something for users to look out for. Fortunately you had the foresight to program in a way to bias the speed so a user can put in 130% of rated speed or whatever to get closer to where they think they might drive.
Phew, glad I fixed it before you reported it Thanks for noticing and taking the time to come here and report.
Travel speeds are no longer from Google's API. Just like Tesla, ABRP uses OSRM (related to MapBox) for backend routing and OpenStreetMap map data. So speeds are determined by speed limits, if known, otherwise a number estimated by road classification and other metadata. I guess the I10 in west Texas lacks proper speed limits in OSM and the estimates for North America are intentionally low. But if you add data to OpenStreetMap it will trickle down to ABRP with some delay!