No, you're assuming the regulation is an end in itself and coming to a tautological. Regulation is essentially saying 'wait, let's factor in these unaccounted-for-but-predictable costs (like repairability)' and you're saying 'but what about the costs of assessing the costs?!'.
Yeah, incorporating some product standards is going to raise the cost of all units sold, which represents a loss to the buyer who hoped to get the product at the cheapest possible price and was either willing to take perfect care of it or tolerate it not being fully repairable if it were broken. But that's OK, because when regulating something the most important person to consider is the median user, noth the clumsiest/unluckiest nor the most skilled/luckiest.
If the median user buys a car expecting to drive it for 5 years, and there's a >50% chance that it will suffer some damage to some major functional component during that time, it would be pretty inefficient for said damage to cost more to fix than the cost of replacing the whole car, unless cars were to become so cheap that it was actually cheaper and better to scrap them - perhaps thanks to some breakthrough in large-scale 3d printing technology or so.
That's what happened with computer keyboards; it used to be that they were fairly expensive and if someone spilled coffee into one it was worth dismantling, cleaning, and re-assembling, but then cheap membrane keyboards assembled with increasing levels of automation brought the price down faster and faster, and today you can get a PC keyboard for something ridiculous like $10. A few fetishists still pay out for old-style IBM Model M, but most people will just replace a damaged keyboard.
But while cars remain relatively expensive and hard to dispose of, it makes sense to minimize the economic losses stemming for a lack of modularity and repairability across the market as a whole, rather than seeking to minimize the costs for the stingiest buyer.