Liking the timeline so far DVB!
Just wanted to throw my two pence into the whole Indian subcontinent war. This period was pretty bad in general for the whole subcontinent, with instability in Indian Punjab along with Afghanistan in chaos and Sri Lanka also beginning to see the beginnings of its own civil war. But with an Afghanistan that does not collapse, there will be major butterflies all around.
So the population of Afghanistan was around 15 million in 1979. Taking a par growth rate of 3.4%, this will give a population of some 19.5-20 million people just before the war kicks off. Pakistan on the other hand during this time had around 100-110 million people, so a population 5 times greater. Taking into account the top of the line (relatively) stuff Afghanistan has compared to Pakistan ITTL, and we can initially see Afghanistan making quick gains. The challenge will then be to keep making gains while holding on to what they have achieved.
Now a key point here is that Sino-Pakistan relations were already strong by this point and an earlier cutting off of American equipment being sold to Pakistan means this relationship would have deepened earlier on by the 80s here. Add to that the fact that the
Karakoram Highway opened in 1978 (something that would still happen here) and you could really be seeing some major Chinese actions also occurring during this whole regional conflict. Now one of the main issues with the war you have presented is that China would not be sitting back and letting Pakistan take so much of a beating without trying to combat this in some sort. Now obviously I reckon the Afghanis would have made cutting off this highway a top priority as it prevents Chinese supplies from entering Pakistan and supporting their war effort, but I think some sort of explanation will be needed as to why China is not getting involved either diplomatically or even launching some attacks of its own. Perhaps the Soviet Union applies some pressure to ensure China stays out of the situation overly, along with the fact that both it and India have nukes even if Pakistan doesn't and that whole situation could turn ugly really fast. This would thus leave China being the one which would thus ensure the survival of a rump Pakistan, more on that in a bit.
Now wars between India and Pakistan have tended to not go on for longer than two months (
the initial war over Kashmir admittedly went on for a year, but during that period India was also ensuring other states forcefully acceded to it, including launching
Operation Polo to capture the Princely State of Hyderabad), with the war which liberated Bangladesh lasting a mere 13 days. Thus, for the war to go on for a year and a half at this point means that India must not be pursuing an aggressive strategy as otherwise a double pincer would mean the war should really be over in three months. This would make some sense as India was struggling economically at the time, and if you add India also getting involved in Sri Lanka like OTL then that would probably also help to explain the length of the conflict. Just keeping the focus on Kashmir should be enough to stretch the war for a long period, with India playing a defensive role along the rest of the frontier for most of the war.
There is also the point to note that Afghanistan would not be able to handle the ambitions it has set out to conquer - the North-West Frontier province at that time alone contained 75% of the hypothetical population of Afghanistan here (around 14-15 million) and Balochistan would just not be possible without Iranian intervention (which I don't see as occurring as they are likely trying to digest their victory over Iraq). Thus they would likely be looking to call it quits after capturing the NWFP, which I agree would take them a while and make the war last longer. So in the end they would see a moderate increase in land but a huge boost in population. India will likely look to capture all of Pakistani Kashmir and finally finish the job, so to speak. The point on the conflict between the Shia and Sunni in Kashmir might be a really interesting aspect to explore too. We might even see a "second partition" where the local Shias begin to kick out the Sunnis which leads to reprisals in Pakistan and thus another population transfer where the Sunnis leave and are replaced by Shias from the south. I could see the war being taken from an Afghan-Pakistan war to an Indo-Pakistan war with Indian offensives to take Lahore and drive to Karachi, with the aim to
capture areas of significant Hindu populations in Sindh and keep them after the war along with Kashmir, in 1989 because it happens to be the Indian general election year lol. This will probably lead to successful offensives and lead to China throwing in the towel and threatening to actively get involved unless all sides come to the diplomatic table. Ironically, America might be the best place to sort this out and it could become an early major foreign policy victory for President Askew. As for the final borders, I expect Afghanistan to gain the NWFP and probably push their borders a bit south into Balochistan while India picks up all of Kashmir and parts of Sindh. This would still leave a sizeable rump Pakistan, although at this point it would likely just become completely dominated by Pakistani Punjab. This nation will likely become completely dependant on China, but will suffer now that there is no land border with China.
Not sure what happens to Pakistan after, maybe it falls to civil war after such a catastrophic defeat. With the remaining significant population of Hindus now in India, Shias will likely become the scapegoat and lead to the forced population transfers I mentioned earlier. Although the war would be a large drain on India, I can see Congress winning quite nicely - especially if butterflies mean Rajiv Gandhi is assassinated in '89, perhaps by Pakistani backed groups. This would mean Rao comes to power in '89 and would likely lead to Indian liberalisation in that year itself (significant because the balance of payments would have likely gotten even worse than it was in '91). Afghanistan would also see a massive surge in popularity for the ruling parties, with the new areas seeing a mix of support for joining Afghanistan and hating the fact they are now in Afghanistan respectively.
Another point to consider would be the equipment needed to fight the war. So likely China will manage to somehow supply Pakistan with its knock-off Soviet arms, while both India and Afghanistan buy loads of arms from the Soviets. And here lies an interesting matter. This might actually give the Soviet economy quite significant short term relief because it would mean cash flowing into their coffers through the selling of equipment and give Gorbachev some breathing space (along with all the money they have saved from decommissioning nukes, not getting involving in Afghanistan, not having ti clean up Chernobyl and now seeing a lot of arms, their economy is probably significantly healthier than what it was compared to OTL). Now the selling of arms matters because we can assume this war will probably be attrition heavy and thus both sides would likely use up a lot of equipment and so there will be a steady stream into Soviet coffers. There is also the fact to consider that the Soviet Union can now build pipelines directly to India to supply them with gas which would be a huge boon for both the Soviets and the Indians.
Sorry for this absolutely large paragraph wall, but hope it helps!