OK, let's do a real world example. I take a drug that's INTENDED to depress my immune system, so I assume I'm at a high risk although there's no data on that yet. My wife isn't at high risk. So the virus is allowed to spread widely, and if I stay home I'm good, right? Oh, no, that's wrong, because to protect me, my wife also has to be quarantined since if she gets it, the odds of her spreading it to me are very high! Now do that for 10s of millions of people, wives, husbands, parents. If their kids get sick, it doesn't matter what dad with diabetes and high blood pressure does because the kids will infect him, or his wife will. And how exactly do I self quarantine in my own home, that I share with my wife? Should I just move into a spare bedroom and have her leave me food three times a day, for six months to a year?
The point is that quarantining 30-50% of the population at higher risk is actually quite impossible. If you're suggesting that as an alternative strategy, it just means you've not spent 5 minutes thinking of the practical difficulties of doing it. Even with an isolated population like a nursing home, it is enormously difficult, because the 40 staff at my mother in law's place have kids, husbands, etc. who work in the real world, and if they get it, the workers do, and then spread it before they are symptomatic. That's why the strategy is a broad based effort to limit the spread, to EVERYONE.