True, the 2012 Giro was extremely close, but was one of the worst raced GTs in the history of racing. I think a lot of the problem has been caused by a couple of races that should have failed massively but didn't.
The 2011 Tour was downright awful for two thirds of it. Completely backloaded, apart from a few seconds collected in to Mûr-de-Brétagne and the farce with the crashes on Mont-des-Alouettes that led to a group crashing inside 3km to go blocking a group that crashed outside 3km to go, doubling their advantage to them, everybody was on their TTT times until over halfway through the race, except for half the GC field crashing out, because the field was nervous because of still having something to protect so everybody was trying to get to the front. The Pyrenees were raced extremely negatively to the point of pure tedium and Jelle Vanendert looking like the reincarnation of Lucien van Impe. With this in mind, when the racing DID begin, Schleck had failed to take advantage of terrain that suited him to the point where his only stylistic option for victory (remember, at this point he was not yet the winner of the 2010 Tour as the appeal was still going on) was the long distance raid. As a result riders who may otherwise have left everything to Galibier were on the rivet early, which meant that when Contador launched his attack the following day in the 110km stage, everything upped and exploded. This mean that the race ended with a series of really good stages that sorted out the GC, everybody left happy because they'd forgotten how boring the start was (much like the 2016 Giro in fact) and ASO thereby concluded that the ideal formula was to backload all the mountains, make the mountain stages really short and stick one TT at the very end.
Likewise, the 2012 Vuelta was born out of a complete folly, the idea that as the Vuelta's mountain stages weren't producing the kind of time gaps expected in the period since Operación Puerto had shorn the Spanish péloton of many of the climbers capable of overcoming the time gaps produced in the comparatively long TT mileage since then, rather than produce
better mountain stages, they'd produce
more mountain stages, but because there were so many stages with these tough finishes, so that the route wasn't too brutal, they'd be relatively short and not include too many mid-stage obstacles. This led to a farcical route with a dozen uphill finishes and barely any time trialling, but with most stages coming down to the final climb. However, due to Contador's ban during 2012, Rodríguez targeting Giro-Vuelta after the Tour route was so TT-biased, and Valverde messing up his peak, and thanks in no small part to only the last 90 mins or so of stages being broadcast in much of the world, the race was a huge success, with GC action nearly every day as the finishes suited at least two of the main protagonists, as well as them being Spain's most recognizable and successful cycling stars. That race had no right to be a success, but it was, and organizers have gone back to that well too often too.
Anyway: the 2017 Tour. As you might expect: it's horrible.
Now, it's not
uniformly horrible. The Mont du Chat stage is perfectly good even with the gap before the final climb (reminds me of the 2010 stage that included the Col de la Madeleine, which was a good stage).
The gap between the batch of climbs and the final climb is longer, but then the climbs before it are harder in the Mont du Chat stage.
The Pyrenees are execrable as usual. Would be much happier (though still not super-happy, at least it would be a change) if the Pau-Peyragudes stage went through Bagnères-de-Bigorre then either Aspin or Hourquette d'Ancizan (probably Aspin from this side, both cat.2), Col du Lançon (cat.2), Azet (1) and then Peyragudes directly - this would only be around 140k, but would have an ending that backs climb into climb well as well as doing something with Peyragudes other than what we've seen every time it hosts. Then the Saint-Gaudens - Foix stage could be freed up to use the Col de Menté (from its harder West side) and Portet d'Aspet before the main climbs of the existing stage (possibly excluding Latrape depending on what ASO want to do climb-wise).
The stages to Rodez and Le Puy-en-Velay seem to have some potential, but once we're into the Alps, I'd prefer to see La Mûre direct to Izoard if we're keeping the MTF there, go over something like Col du Noyer early (cat.1) to get a strong break, down via Col du Festre to Gap, Col de Manse (cat.2), Col de Pontis (2), then Vars-Izoard. Then Briançon to Serre-Chevalier, if the Lautaret tunnel is open then Lautaret - Glandon - Mollard - Télégraphe-Galibier is perhaps best, if not, Montgenèvre - Mont-Cenis (preferably via the steep and inconsistent Italian eastern side through Moncenisio itself) - Télégraphe-Galibier.
I think the less said about Liège and Planche des Belles Filles the better, really.
So, summary:
- Make Liège a legitimately hilly stage, don't be scared of 2010
- Make the run-in to PdBF a bit more interesting, congrats for sorting the pretenders from the contenders in week 1, but if it has to be that climb then at least try to thin the bunch before it
- Make the Périgueux-Bergerac stage an ITT down the N21 (45-50km). That's the only way that the Marseille TT being so short can be acceptable, with a 13km one to start and a proper length one in the middle, like the short TT that finished the 2009 Giro after the 60k Cinque Terre one for example.
- There must be more than one stage design available to a ski station situated where Peyragudes is?
- Foix stage is a joke
- Disappointing weekend stages on penultimate weekend yet again
- Swap the hosts around for the Alpine stages and you might get better stages, certainly less predictable ones
- Very disappointing final mountain stage
- 23km final TT is ridiculous when there hasn't been a proper one before it.