All these points are just nicely written bits of my own long held views. I've always known that manufacturers build into their vehicles a "useable life". They don't want their cars lasting too long - they need people to keep buying their new models. Millions of other products are built like this too - it's a whole science within (or probably more likely separate from) their engineering departments to create a product that fails at a certain point in it's life. But not appear to do it too obviously. To hide a programmed wear item away in the torque converter is a sly but effective way of doing just that.
But nobody told Westinghouse that in the 1970s and early 80s. I have both a Westinghouse fridge and stove, built in the 70s and 80s respectively. Neither will die, so I keep using them and respect the engineering of the day. The fridge in particular, which was the top model at the time has a heap of different features and options - and all still working. In fact it works so well that I never have it above setting 2 (out of 6).
I don’t think that there is planned obsolescence as such, or rather if there is, it’s a byproduct of meeting a market price. So the motivation is to be competitive in the new product market.
For example, some of the all-wheel-drive cars use clutch packs and electronics to distribute the torque rather than gearboxes or transfer cases. There are two advantages… the electronics makes it work better (from a driver’s perspective) and it is cheaper to make and lower weight (than the traditional gear arrangement). So this gives the manufacturer a marketing advantage (our vehicle with superior driving performance is available cheaper). The trade off is that clutch packs will wear out (so is less reliable than the geared arrangement).
So, how does the manufacturer deal with the clutch-pack arrangement and determine how long it will last. Quite easily. All they need to do is run a few test vehicles and carefully weigh the clutch packs before and after a test run. They can then calculate the amount of material worn off the clutch pack per kilometre and then calculate how long they’ll last. My guess if is the calculated life is 160,000 or more kilometres, then it is good to go.
So the NVH reason for allowing the torque converter to slip would have been calculate and the optimum point between acceptable life and acceptable performance (low NVH) would have been calculated.
In the end, manufacturers are selling their cars to ‘new car buyers’. These buyers mostly buy new cars regularly so are only looking for reliable motoring for the length of their ownership (or the length of their lease) - say, 3 to 5 years - plus enough remaining reliability to preserve a reasonable residual value. The people who really care about longer term reliability are the secondhand car buyers, but they are essentially picking over the bones of the new car market (and are not a driving influence on the new car market or new car design parameters).
As an aside, be careful of old fridges. Yes, they are incredibly reliable and will probably outlive cockroaches, but they are also very inefficient. A few years ago I replaced a 25 year-old Hoover fridge as I remodelled our kitchen. The old fridge was still working perfectly (and had only broke down once - a defrost timer failed). Before replacing it, I decided to do some energy measurements on it. I don’t have the hard numbers at hand, but I worked out that the new fridge (costing about $2,000) would pay for itself after about three years - just in the energy savings (or running costs) between the old and new fridge.
I usually caution people against putting the out fridge in the shed (or wherever). If they really need another fridge, then go out and buy a new one rather than use the old one as a ‘drinks fridge’.