Might be interesting to some folks.
What I do is a benchmark test on an oil based turbo trainer, cyclops. It's a ramp down with intervals.
Here is a group of intervals at a specific load, 1 min hard, 1 min at approx 50-70% of the max load by reducing my cadence to 70 rpm. I adjust the load and cadence by changing gear, ie change down to increase cadence and reduce load. I'm no spring chicken and my max HR is around 165. So these could be expressed in % if you wanted to like for like comparison.
Last night I did
Reps MPH Cadence Avg HR Max HR
4 28 89 130 148
7 27 92 134 150
8 26 95 130 147
Followed by 5 mins at 25 mph none stop 97 rpm Avg HR = 145 Max HR = 151
Then 5 mins easy spinning cool down (100+ rpm).
The whole thing took less than an hour.
What I can see from a year ago is that the Avg and Max have decreased for the same load. I've also increased the number of reps at a specific load and the load I can take now compared to 2 years ago is significantly greater.
You can also see a bit of thermal effect as I get hot when the max hr goes from 148 to 150 on slightly lower load. The continuous 5 mins is a higher avg because there is no easy 1 min section in between, but the max is higher too due to the continous load and me getting hotter sweating buckets.
The 26 mph is too low to provide strength training,the HR has come down, but since cycling is about muscular endurance I know from the pain in my legs on end of that secion it is doing something, lactate tolerance possibly.
After a test like this my HR stays elevated for some hours and the next day I can feel the effect in two ways, I don't have the ability to max out my HR and peddling feels a lot easier - I guess it is down to muscle memory effects.
None of this would be possible if I had not kept a record of what I was doing.
HTH