So the winner in 2012 does not include 64 minutes at 6.4W/kg (!!) from the final TT.
Can you please confirm whether Lance's placing in Tours remains in play, or if everyone finishing after him was shifted up one? In particular it looks like Wiggo (4th 2009 including Lance) - in this limited dataset - generated less power on the climbs in 2012 (1st) than 2009 (4th), despite (allegedly) weighing less in 2012.
Because climbing is predominantly overcoming gravity, which remains a constant for all riders. TT, being aerodynamically dependent, means you have to estimate the frontal surface area, and wind on the day, even to the closeness of following and leading vehicles, can have a dramatic impact.
Wiggo admitted his final TT power - that's how I know what it was.
Too many variables in the TT mean estimating it accurately is too difficult.
One other point not mentioned above is the propensity of the favourites to look at the TT as a way of gaining time over the pure mountain goats,
If you know you can put time into some of your rivals in the TT then you are more likely to work on it and work on position and power, But they may also want to take advantage of a boost in other ways, While the rest of the pack are taking a day off and not topping up the tank with whatever goodies are on offer from the sweet cabinet, the good TT guys may be jacking up the program the night before to ensure the extra boost on TT day to drive home the advantage as much as possible.
So while we know most of the top guys will be going full genius on the MTF stages and the big selection stages, the pure climbers may not be on the TT stages bringing another variable to the TT data and making it less useful as an indicator.