Hello 
Just a few ideas :
(1) It could be great to plot the ascent speed curve on the activity page. A curve smoothed on 30seconds.
(2) You could use the new page “compare” that you recently developed to compare the average ascent speed function of time (a kind of critical VAM) over seasons and gradients:
I have added (1) to the todo list.
(2) is a bit trickier to do. That would likely need to be pre-computed like the power vs HR data. Is this a running thing? I have a bunch of running stuff on the todo list.
I am generally curious about these quantities. Maybe this is because I’ve always lived in places with barely any hills. @Hugo_Joudrier, how do you use the ascent speed and VAM?
Not sure about running, but VAM is just another way to look at W/kg, so it’s generally not that informative in my opinion.
Showing W/kg vs cadence (or torque vs angular velocity) vs grade would be interesting imho.
Keep in mind that there should be a check that source data is coming from a device with altimeter and not from gps data, since the latter introduces a lot of noise in the computation.
So does your head unit, if signal (reception) is poor.
I use it to have an idea of the average power (W/Kg) when I do not have power meter. The situation occurs when I ride a bike not equipped with any power meter and when I run. I also use it to compare bikes equipped of power meter.
Before I invest in a power meter I was riding using VAM (smoothed over 30sec) to have an idea of the intensity of the effort I was doing. Now I have a power meter then I do not use VAM so much, now I use VAM to compare my recent rides with previous seasons (when I was not using power meter) and for running.
You’re correct. I should have added that a second qualifier for data to be usable is the presence of a speed sensor.
Once you have ground speed and barometric altimeter data the only thing missing is air speed, and that’s another can of worms 
True, but in general less of an influence when climbing. Obviously more so on steeper grades - if you’re on relatively flatter straights and face a crippling head wind, then it does count.
Although, I recall climbing the Izoard, with a howling head wind op to Brunissard, where the grades are > 10% and I was glad my team car was giving me some shelter 
btw, if you want projected estimates of your ascent speed and time…working from your FTP (or your actual climbing power) you will find it here: https://www.cyclingapps.net/calculators/climb-power-calculator/ (FTP power) https://www.cyclingapps.net/calculators/climb-speed-calculator/ (fixed power)
Thanks for the links, I was aware of those spreadsheets.
I have a couple python script that actually break the climb in “gradient brackets” instead of one single elevation step, and allows to set an ideal power for each grade (that’s why I was asking for a way to chart historical power/vs/grade data) and estimate climbing time. Most of the climbs in my area tend to have “steps”, steep/flattish/steep
You often speak of a can of worms - I think you found another.
On a related note: most profiles of climbs are equally handicapped, as they are usually divided in 1km (or 500m when shorter) stretches with averages. These don’t mean shit.