This sounds like an argument for the 'proposals' (to give them far too grand a term) in the Great Barrington declaration.
There are fundamental issues with shielding the elderly/vulnerable and letting the virus run through the rest of the population:
-Those vulnerable people have support networks. To maintain that requires all those carers to be effectively isolated too, because there's too many of them to test regularly/rapidly. Care homes, etc would need to be sealed off from the outside world, with all staff within. As seen in March/April, once the virus gets into a carehome it can do untold damage, and this will be even harder to prevent when the virus is widespread outside of the home. Food, medicines, medics, post, etc will still need to come in. Visiting carers would have to be isolated from anyone they aren't supporting. And that's not to mention replacing the childcare, etc and other functions of these vulnerable people forced to shield.
-70+ would appear to be the likely cut-off for shielding in the event of a "let the virus spread" policy. That encompasses nearly 8m people, who would be forced to isolate for an indefinite period, which could easily run to 12 months from now. Whilst a month-long lockdown undoubtedly has health impacts, I cannot see that locking 67 million people down for one month is as bad as locking the most-vulnerable 8 million down for a year.
-A nationwide lockdown helps make it possible for the NHS to handle health problems beyond COVID. Letting the virus spread will put a great burden on the NHS. The median age of someone in intensive care due to COVID is only ~60, so allowing the virus to run through the population is still going to result in a huge number of hospital admissions for COVID, and dilute the quality of healthcare which can be provided for COVID and non-COVID patients. "Protect the NHS" was an accurate part of the government's slogans.
-There's no guarantee that having the virus prevents you from catching it again. We could let it move through the "healthy" population now, only for them to be struck down again in six months time. Letting the virus run through the low-risk population is a gamble that herd immunity can be achieved through infection, and one that would do great damage if incorrect.
-"Long COVID" is still an ill-defined medical condition, but one that may be present in 5% of cases (
https://covid.joinzoe.com/post/long-covid). Letting COVID rip through 60 million people may therefore leave up to 3 million people with long-term health conditions which will place a giant burden on the NHS, and the economy if these people cannot return to work. As with the "short-term" disease, time has allowed scientists and clinicians to study and develop treatments and medication which greatly improve the prognosis. More time is needed to study these long-term conditions before it can be even remotely ethical to allow a great deal more people to be infected.
-When the virus is rife, people change their habits. Even if restaurants are open, far fewer people will go whilst the virus is so prevalent. Businesses may remain open, but not be viable and close anyway. Even if we have to have 4, month-long lockdowns a year until a vaccine is developed, that gives businesses 75% of the year with "normal" footfall, and makes it much easier to plan the support needed for busineses/employees.
-When a vaccine is first available, it will initially be in short supply. With COVID at low levels, it's much easier to target vaccination at those people/areas that need it most/it will have the most benefit to. With a widespread outbreak it's much harder to bring it back under control until a very high level of vaccination is achieved.