I suspect that Google doesn’t account for anything. If I had to guess, I’d say it assumes a certain pace as the normal walking pace, and leaves it at that. Elevation, traffic lights, fatigue, and any other factors that might affect it be damned. So its time estimations for a hiking route are going to be a lot worse than its regular walking estimations. And all that’s before you account for, as you say, individual differences.

Fwiw most people I’ve seen recommend switching from Organic Maps to CoMaps. The latter is a fork of the former, made because OM was being run in a very untransparent way, including (but far from limited to) the discreet use of closed-source code.

I don’t hike, but my experience with regular walking directions is that Google is regularly wrong…because it overestimates how long a trip will take.

Its directions for driving and public transport (at least where it has cooperation of the local public transport agency, as it does where I live) are excellent.

Cycling directions are the only one where I’d say a normal person’s experience is likely to be a bad one, with it regularly taking you on dangerous busy roads rather than nearby quieter streets or bike paths.

So I don’t find it particularly surprising at all that someone who isn’t a regular hiker would make the mistake of relying on it.

Yeah that’s an insane amount. Nearly 1 in every 50 joules generated is currently going to generative AI. That’s incredible for a single purpose.

Wikipedia has entirely different versions in different languages. This page is on the English Wikipedia.

Yeah that seems to be how Wikipedia does things. It tripped me up earlier today on a different page. Bizarre. Just go left to right, top to bottom. Reading order.