it's depending on the given data: most accurate power , actual bodyweight, actual maxHR , actual restHR (for sure, the 3 most important is the accurate power, weight, and maxHR). lots of people leave the maxHR on an old high maxHR value (when the fitness is very low) and this can cause high estimated vo2max value in some cases as he/she is starting to train seriously (if other datas are correct). but what i can tell the x40 series has more accurate algorythm, i had 830 also, and i could fool it indoor with special ERG sessions (i had 78-80 VO2Max with that but i knew it is extreme high), with 840 (with same refeference accurate powermeters, P2M NGs, and other accurate datas) i had "only" 69-71. but probably my maxHR value was set too high in 830 then... the x40 series can adjust your maxHR up or down by an own Firstbeat calculation method (if the maxHR detection is enabled, this value comes from a HRV derivation and not from the highest HR value from the activity) so probably this is one of the reason why the x40 series more accurate
but you can do a "labtest" by your own also outside, do an all-out 6min effort on a steady hill, and here is the calculation:
VO2max=(6min avgPower * 10.8 / weight)+7 , in my case it was the situation (391avgW*10.8/68kg)+7 = 69.1 so the Garmin estimation was right in my case (my theorically achievable best 6minavgPower is 408w by power curve, so my theoritically VO2MAx is/was 71.8)
what i noticed, lots of people enter too low bodyweight (or leave it on a low value but gained muscle/weight ) and/or the maxHR is not accurate (or certain circumstances raised the HR extremely during a ride, i mean coffein intake and/or heat and and/or not drinking enough during a training and that measured maxHR is a unique high value and not the regular one) or the used power meter is overmeasuring the power, so there are lots of "fake" high Garmin VO2Max screenshots on the Strava or other places. the 6min all-out test doesnt lie!
