I am having hard time understanding your calculation. How did you arrive to the conclusion that 160 kg of plastic getting 'disposed' is equivalent to only 304 kg of oil? Are you ignoring the cost of processing and manufacturing the part and only accounting for raw materials?

I will assume you are arguing in good faith and just never owned an old car. Here is approximately how bathtub curve works in practice.

Year 1 to 3 is warranty. Anything that happens in this period is fixed by the dealership at no or minimal cost to the owner. These early failures are cost to the manufacturer and not car owner. Depreciation is steepest during this period.

Year 3 to 6 is 'optimal', failures are rare and usually minor. Car depreciation is average.

Year 6 to 8 is when wear and tear items have to be replaced. New tires, brake job, some suspension components are to be expected. Big failures are possible, but not typical. Depreciation is below average, as long as you sell before 100K.

Year 8 to 12. You start seeing major component fail. Issues like transmission failures, emission system repairs, major suspension components. It is unusual, but possible to have the car mechanically totaled during this period. Typically luxury vehicles get totaled in this range, as they depreciate most while costs to repair are highest.

Year 13 to 15.This is period we are discussing. Usually it is 3d of 4th owner with typical mileage of 150K+. By year 13 vehicle usually depreciated 80%+, so a typical non-luxury vehicle will be worth 10K or less. Serious issue, like transmission failure would not mechanically total it, but cascade of issue would. Additionally, the vehicle no longer get fixed to 'everything works' standard, but to much lower 'does it drive' standard. This results in accumulation of failures due to deferred work that results in it getting mechanically totaled when a major failure happens.

TL;DR You are wrong in your interpretation of bathtub curve as a rapid cascade of failures leading to the vehicle getting totaled. In practice, this does not happen. What happens is frequency of failures increases from about 1 per year for 8 year old car to 3-4 per year for 15 year old car. The accumulation of unaddressed failures is what results in the car getting written off. This accumulation is the direct result of cost of repairs (i.e., more expensive is to fix something, more likely it will go unaddressed unless it prevents the car from getting driven).

This does not follow. One of the circumstances when old cars get scrapped is when a major fails component fails and the cost to repair exceeds the value of the car (i.e., mechanically totaled). As cars depreciate with time and use, cheaper the repairs, older the point where it would get mechanically totaled.

Absolutely, categorically no. No emission compliance equipment ever improved the car performance, longevity or made it cheaper. The closest would be fuel injection, but early mechanical injection systems were a nightmare to maintain, it is only much later when electronically controlled injection was developed by Bosch did things improve over carbureted engines.

Typical development cycle of emission equipment is: 1) unreliable expensive crap that nobody wants, 2) unreliable inexpensive crap that nobody wants, 3) somewhat reliable inexpensive crap that nobody cares about.

Using this logic, then tailpipe emissions are also irrelevant.

Go visit a junk yard. It is ecological disaster all around. We only pretend that cars get recycled, the reality that plastic, rubber, glass are not.

The right to human review and override needs to become the next Amendment.

In context of my argument we don't have these additional considerations as it is direct comparison. My argument is very simple - if it costs too much to repair then people will scrap the car earlier than they otherwise would and this in turn results in higher lifetime emissions. Therefore, when optimizing for tailpipe emissions we need to also take into account effects of added complexity on repairability.

To demonstrate by example. If Car A and Car B both have the same bathtub curve, but Car B is twice as expensive to repair, then Car B will get scrapped earlier than Car A. Therefore, everything else being equal, Car A will have lower lifetime emissions.

Citation required for a full cycle (that is includes accounting for non-recyclable materials like plastics) emissions.

Anecdata: Grandma down the street from me driving 90s-era Camry.

It is absurd to compare an old car to intentionally malfunctioning diesel. As to rolling coal, this would be illegal pretty much anywhere in EU, so you are already have your zone.

You missed my point. Each car has a lifetime emissions associated with manufacturing and disposal. Shortening useful life results in overall emission increase due to that. So more complex cars that are not getting fixed due to prohibitive costs will emit more due to that. You can't over-optimize for the tailpipe emissions.

I imagine not that many. Especially considering Citroen is rather unreliable.

Gasoline tax is regressive.

Funny how that also works with emissions. Cars that are new right now are much harder to fix due to all complex emission equipment, nobody going to pay prohibitive costs to fix them once they get old so they will pollute more (or get scrapped early) increasing the overall emissions. Meanwhile 60s VW Bug and 80s Toyota Hilux will still be around.

This is being forced by definition for any value of X != Y.

This is really twisted way to misrepresent the issue of freedom of movement. You can use the same broken logic to justify killing people, because everyone consumes some amount of common resources.