Thanks for that. I knew there was a more robust way to normalize for HR vs. intensity. I hadn't gotten around to looking it up.

Interestingly (to me), I put in my own data for my two runs yesterday:

run1 = VERY easy steady-state run (top of z1) outside,
run2 = mile repeats on the TM (WU, 2x1m (2m), CD)

I also put in the data from run2 for the warmup, and the mile repeats individually, as well as the entire run "averages". All sections of run2 (warmup, mile1, mile2, and entire run) agreed on my VO2max, exactly.

The easy run estimated my VO2max a little higher (+4 ml/kg/min), but it was also outside in the cold (~45F), vs. run2 on the treadmill (~70F). If I bump the easy HR up to what I would expect on the TM (147 -> 151)...it agrees with everything else.

I also put in a recent 2x20 bike, which estimated as -4 from VO2max(run).

My garmin 920 estimates my run VO2max about 4 ml/kg/min lower, than your formula.

ETA: I also went back and plugged in my runs over the 100/100 challenge. Starting with the Monday before, and every Monday since. Which shows an improvement in estimated VO2max(run) of just over 10% (beginning to now), at an improvement rate of roughly 1.5 ml/kg/min per week in a very linear fashion (not a lot of random variation).