first conclusions without knowing the exact profiles per stage:
-at least it is a very odd route for tour standards, how different it is compared to the routes at the beginning of this century.
-a lot of tricky stages, there can action from stage 2 till 20
-interesting first week, with quite some variation, soft mountain finish (merlette), hard mountain finish (lusette), hilly finish (Privas), flat flat (laveur), flat but not flat flat (nice, sisteron) and downhill finishes (pyrenees), hard medium mountain (nice).
-no 2nd week of boredom after starting in the south (remember tour 2009)
-no parade of ASO 'classics' (no tourmalet, no Aubisque, no Alpe d'Huez, no Galibier, no croix de fer, no glandon (although, I wouldn't mind if they had put it in the stage to loze))
-lot of interesting new climbs/ not used for a while
-not backloaded, but difficulties spread out over the whole course
-barely real sprint stages, or they have to work for it
-tricky finals in Privas, Sarran and Lyon, both seems to have hills in the final. Lyon even seems to have col de beal in the stage.
-last mountain stage not being a tourmalet-hautacam kind of stage (the kind of stage that blocked the whole last massive, without being interesting itself)
-proper echolon stage
-sterrato (although more a bit of a gimmick than really making the stage more difficult)
less positive I'm about
-2nd pyrenees stage going soudet downhill to arette, instead of the eastern downhill straight to marie-blanque. It really ruins a bit the idea of having the Soudet in the stage.
-I think they wasted a bit the stage to villars-de-lans. Those mountains are really to soft/ not steep enough to see any action. Mont-Noir would have been nice, as some other possibilities as well.
-Both queen stages are missing a bit of difficulty. Glandon or Champ-Laurent before madeilene (would have made it a proper queen stagen and reduced the risk favourites waiting for the very steep last 5 kilometers of loze) and Bisanne instead of Saisses would have made both stages much better. The combo Saisses-Aravis is probably a bit soft.
-Of course a 25-35 km flat time trail somewhere in the second week would have been great, to see whom of Froome, Thomas, Roglic and Dumoulin is the best time-trailist
-risk of planche des belles filles blocking the race a bit
-no time trail in the beginning, could means small gaps between the contenders and could lead to less willingness to take risk/ or at least attack. Giro 2012 alarm.
-no endurance stage(s) to test the biggest tanks in the peloton. ASO is really going Vuelta in that sense.