In real war situations, it is very difficult to maneuver tanks without exposing them to enemy action. The text books say you should use all your tanks in one place and drive them all in the same direction. But in the real world, the terrain is uneven, and if you put all your tanks in one place, bottlenecks can develop - especially if one or more tanks become disabled for some reason. Another recurring problem is if you have to crest in the middle of the proposed line of attack, putting all your tanks in one place means you have an entire regiment size force performing cresting maneuvers in the exact same place at more/less the same time. The risks inherent in that are very great. I think it may be difficult in practice to use tanks in a single punch like the theories of Guderian et al. recommend and a battlefield commander has to do whatever the situation calls for.
Presently - one is faced with two highly agile adversary technologies - advances in multispectral imaging and anti-tank missiles. Those thermal cameras on drones make it all too easy to spot a build up, and if the build up is in place before enemy has been cleared from the sky, the entire tank force will be killed. Even if by some miracle the enemy does not kill the amassed force from the air, the mere awareness of its existence will allow the enemy to beef up their posture and prepare the defending force for the arrival of the tanks and set up a kill zone where possible.
To give an example in aksai chin area, you could imagine that a Chinese tank force will be able to cut their way from the western shores of Spangur lake to the Chusul airstrip, but long before they get there - they will be spotted by the defenders and corrective action could be launched to create a kill zone along the Tsaka La road.
On the western side, if the Azm-e-Nau exercises are anything to go by - the Pakistanis are going to avoid the "Houbara Run" situation by letting loose a bunch of chickens in the houbara's place. Their belief is that if the IA strike corps comes across the border, it will become preoccupied with the chickens and sufficiently distracted for the Pak Fauj to maneuver its reserves without being totally destroyed. Critical to this is the ability of PA/SSG ground intelligence units/LRRPs to pick out the exact location of the IA strike formation. In 2001, rumor has it - the Americans came by with the right remote sensing technology - when the SSG/LRRPs failed to make a concrete determination. That changed the game, and it is rumored that ABV had to visibly pull Gen. Kapil Vij and his boys back. Apparently Gen. Padmanabhan was angered enough to write a fiction novel.
I think the CSD was an effective way to put an end to the Pakistani ideas of offensive defense. I feel Azm-e-Nau is just Pakistani-speak for "Defensive defence carried out in a diligently defensive way". As one ladder of escalation ended in a Pakistani use of a nuclear weapon in a battlefield - against an attacking Indian formation, CSD has ensured that that attack will take place on Pakistan soil.
Now there are new technologies like BAE's Adaptiv camouflage coming on the market and they could improve the individual tank's chances of survival in a combat situation, but in the bigger picture I feel all these technologies are simply adding to the weight of the tank. Along with all the armor, the tank will now need more fuel to operate and that means a greater demand on the supply lines.