Originally Posted by TMonk

I do step 3) (the VO2 max interval), but admittedly I wouldn't call it "all out", as I typically ride it at ~10% more power than what I aim to hold during the 20'. I'm equally lazy about the rest of the protocol.

The utility in any FTP test (emphasis on that first letter) is that it helps establishes meaningful training zones. To that end, I think that consistency in an athlete's methodology (and result) is in ways more useful than the absolute accuracy of the calculated threshold power. As previously discussed, a true threshold effort itself has limited utility in racing. Empirically I seem to be able to attack and roll breaks (or TTT's) with guys that greatly outclass me from a threshold perspective, but the former plays to my strengths more.

I was thinking about it during my ride yesterday, and if anything, an overestimate of one's FTP is going to make their shorter intervals harder than what they need to be. At the time I thought, well that's better than setting your zones too low and not getting large gains from your efforts, but obviously there is some risk associated with that due to burnouot etc. For me personally, I know that I have a pursuiter's profile and have good anaerobic recovery, so setting the bar a little higher (relative to my true FTP) is probably what I should be doing anyways for efforts in those durations.

tl;dr, it's important to remember the "F" in FTP and to be consistent in one's test method.