I still think accelerating is over rated. The quickest way to the top of a mountain is to ride a steady speed. As demonstrated by lots of people on Ventoux, you don't have to attack to drop people.
I can see where you are coming from with this and if it were say a time trial I would probably agree that the fastest way up the mountain would be a steady climb. However, we are actually talking about a race here where other factors are involved.
Consider this; If Froome, Wiggins or Contador for instance were the fastest up any particular mountain, why would they benefit from having a team mate pace them for part or all of the climb? Surely they would all just start off at the bottom at whatever pace they could manage and the fastest rider would then just outstrip all the rest?
Obviously that isn't the case and witness Froome's three attempts to drop Quintana each time sprinting out of the saddle I think. On the first two occasions Quintana managed to match Froome's acceleration to get back on to his back wheel and then just resume at Froome's pace.
On the third attempt Quintana couldn't match Froome's acceleration and just had to settle down to the pace he could manage to the top while Froome rode away. Froome also then settled down once more and rode to the top to take 30 seconds out of Quintana.
Do you think Froome could have achieved a similar result if he had just cruised up to Quintana initially and then carried on at a similar pace? You may argue that Quintana had weakened himself by the earlier accelerations that he did to break away from the pack and if he had just maintained his optimum pace all the way up the result may have been very similar.
I think you would probably be right in that assumption but obviously this is a race not a time trial where tactics and positioning are all important, mentally as well as physically as deviant has asserted.