Comparing times for Alpe d'Huez is always a big problem. On flat road, you want heavy guys who have a lot of power in relation to their front surface. For climbing, you want tiny midgets.

This means that the amount of road before the climb affects the speed of the climb a lot. And unless there is contention for the polka-dot jersey, it might not be meaningful to risk blowing a gasket up the climb - especially if the stage doesn't end at the summit. A number of less-than-good climbers have excellent bike control and can take in many minutes on the downhill.

Another thing is what the rules allows, and the material revolution. It helps a lot if you are allowed to switch to a very uncomfortable but extremely light bike optimized for the climb, or if you have to use the same bike you would also have to optimize for 150km of more or less flat racing. Some years, you have been allowed to remove the helmet the last part of climbs. Cooking your head does significantly affect your performance.

In the end, you can't compare times between different races like you can compare the time around Boston Marathon or 3000m hurdles. You don't suffer unless you have something to gain, which is also the reason why some riders who are not so good in the mountains can sometimes score huge amounts of points - no one bothers to hunt them down, and no one is interested in taking any risks. After all, you can win a number of minutes by being a fast climber, but you can lose half an hour by overestimating your capacity...

I think it is wrong to blame TdF. That is like blaming the banks for the existence of bank robbers. The TdF organisation has spent a huge amount of time on making the race clean. The publicity loss of the Landis incident costed tremendously. Both for TdF and for the sport in general. The TdF organisation staff would gladly sell their mothers if they could catch a cheater in time, instead of having the cheater getting caught after having received the final yellow jersey in Paris.