For the purposes of this post, I will assume that the Manmohan Singh administration is not completely packed with traitors from top to bottom; difficult as it is to ignore a proven and incontrovertible political reality, here goes.
A few things to consider:
1) 3:1 ratio. This is a magic number that was thrown around by many "experts" during Parakram, after 26/11 and so on. The idea is that, to make consequential territorial gains, the attacker has to create and sustain a 3:1 advantage in firepower, manpower etc. (including force-multipliers) along a substantial section of the border. This will enable the attacker to grind down the border by attrition.
So my point is: what is true for Pakistan with respect to India, must be true for India with respect to China also no? PLA is not going to make any consequential territorial gains unless they can build up and maintain a 3:1 force advantage at chosen points along the ICB (India-China Border.) If the IA can hold down this ratio to 3:2 or even 2:1, PLA can't make significant inroads.
We know that PLA has a far better infrastructure to rely on for deployment. Yet, I find it difficult to believe the IA's logistics are so incompetent as to be unable to prevent the buildup and maintenance of such a huge ratio at any point along the border.
So ideally, at the first rung of the escalation ladder, we have Indian and Chinese troops deployed in all theatres of activity but with the Chinese never enjoying the 3:1 ratio at any point. Artillery and rocket duels take place, skirmishes occur, but PLA cannot get anywhere. IA can hold most territory, and regain whatever is lost in isolated Chinese thrusts.
2) The second rung of the escalation ladder involves what may need to be done in order to prevent PLA from gaining and maintaining a 3:1 advantage at any point along the LAC/ICB. A major force-multiplier for the IA is the IAF. Two major disadvantages for the PLA are (a) long supply lines vulnerable to interdiction, that must be held open if they have any hope of maintaining 3:1 advantage on the LAC/ICB. (b) high-altitude airfields in Tibet decrease the payload of flights that have to take off so far above sea level.
So the key question is: can India prevent China from enjoying 3:1 advantage along the border, WITHOUT using the IAF to hit logistical nuclei and arteries deep inside Chengdu MR?
Can we fight a "purely defensive" air war, as in Kargil, never crossing the LAC/ICB to any significant extent, and yet prevent PLA from achieving 3:1 advantage?
I ask this because, if we do use IAF in this manner we are giving an opening for China to escalate to its own second rung of the escalation ladder. This is the use of thousands MRBMs deployed in Tibet to hit a large swath of military, logistics, economic and possibly civilian targets all across north and northeast India. At this point we start taking economic hits at a higher order of magnitude than what would be required to merely sustain the IA presence along the LAC/ICB.
It is my assessment... based only on what I have read in BRF, mainly... that even at this second level of escalation, PLA in general would not be able to build up a 3:1 advantage at many points along the LAC/ICB. However, given severe infrastructural damage that their missile strikes will be able to inflict, they MAY be able to achieve a 3:1 advantage at one or very few select points along the LAC/ICB, from where they will launch the next level of escalation in the hope of ending the war in their favour.
3) The third level of escalation will probably be represented by a swift, massive Chinese offensive to seize a substantial chunk of Indian territory, such as a city (Tawang, Leh, Gangtok, or Itanagar most likely... deeper targets such as Gauhati or Siliguri possibly) and end the war on their terms. I believe the PLA might attempt to recapitulate the Russian victory over Georgia (2008) where, once the Russians and their allied militias had seized the city of Tskhinvali, it was all over but the hand-wringing and Georgia virtually had to give up any claims to Abkhazia. http://en.wikipedia.org/wiki/Russia-Georgia_war
Of course, India is not Georgia, but the morale shock of losing a high-profile target like a city, combined with the increasing economic stress of (a) deployment maintenance (b) losses to Chinese missiles across N/NE India might destroy the Indian will to continue the fight and force the GOI to sue for peace on China's terms.
At this point India has very few options. There is no chance of seizing a Chinese city of comparable size; we are simply not equipped to mount a drive to Lhasa even at the opening stages of a war let alone after having exhausted ourselves to the point where PRC has taken an Indian city.
i) We can dig in and hang on, hemorrhaging wealth, for a sustained war of attrition and covert ops to unseat the Chinese from the city they grab. Even an "asymmetric" war so to speak.
ii) We can attempt to expand our airstrikes to Chinese cities within range. Lanzhou, Chengdu, Kunming possibly though not likely. However, if there has been a significant air war by this point in the game, we may not have the assets to sustain any such campaign.
iii) We can gamble very high stakes, and try to make a credible show of escalating to nuclear exchange if the Chinese don't vacate whatever city they have captured.
Most likely we will do none of these, but sue for peace on very humiliating terms.
4) It is therefore of utmost importance that the Chinese never manage to grab and hold on to any substantial chunk of Indian territory, especially a city, for any significant amount of time. In terms of a time scale I don't think India would be prepared to fight this war for longer than a month, possibly two.
However, China will begin to look less indefatigable, more vulnerable, the longer the war goes on and no significant gains can be claimed. Consider the China-Vietnam conflict http://en.wikipedia.org/wiki/Sino-Vietnamese_War
The Chinese were able to capture the heights of Lang Son, virtually surrounding the capital of Hanoi. Their aim was to try and force the Viets to draw in VA units that were then deployed in Cambodia, where they were maintaining security after having deposed a pro-Beijing dictatorship.
The Viets saw through the Chinese plan after the Russians shared satellite data with them regarding the PLA deployment. They did not come out and fight any major engagements with the PLA, but used guerilla tactics very effectively, killing over 70,000 Chinese. Meanwhile the deployment in Cambodia was kept intact, and major conventional units (upto 300,000 VA troops) concentrated for the defense of Hanoi rather than trying to take back the Lang Son heights. PLA was hurting from the guerilla hits, and realized they did not have the strength to ensure a victory in Hanoi without humiliating losses... so they declared "lesson has been taught, road to Hanoi is open", turned tail and left.
The reinforcement of cities and towns will be critical, more so than the prevention of wilderness land grabs by the PLA. We should not spread ourselves thin trying to defend every square inch of wilderness terrain (like the TSPA in Bangladesh) but ensure that the Chinese never take a populated area; meanwhile, guerilla tactics can be used to unsettle Chinese deployments on wilderness terrain within Indian territory.
The question is not just how to make the war more costly for China, but how to dig in for a long fight that will be least costly to ourselves. IF we can dig in and hang on, time is definitely on our side; and if we do it without overt escalation on anticipated rungs of the ladder, using asymmetric warfare wherever possible, China will find itself in a position where it can either escalate to a new order of conflict (large-scale missile/air strikes on Indian cities, etc.) that risks nuclear exchange, or it has to pull up stakes and go home. Loss of face will mount every day that it sits on Indian territory taking losses while not making any significant gain.
This is the point that we must push the Chinese to.
5) At the present time I don't think there will be major conventional involvement from Pakistan, simply because the US is in Afghanistan and (for their own interests) will GUBO the Pakis into avoiding any significant conventional misadventure.
However, there is no doubt the Pakis will do everything they can to give China asymmetric warfare support (upto and including Kargil-type grabs if they can.) They will activate all their proxies in J&K to create chaos in the IA's rear, sow FUD, disrupt supply lines such as the Srinagar-Leh highway etc. This will be no more than a headache for us to deal with in the larger scheme of things, but we should anticipate that it will be there.