In principle, this is a fine idea, but where it falls down is the practicalities of protecting the at risk group, Depending on the definition of at risk you fancy using, you're talking about protecting between 15 and 30% of the population. My stab at it would be 25%.
So you have to pretty much isolate that 25% of the population until virtually everyone else has been infected, which assumes infection = immunity, which, to be fair, in the short to medium term is probably right.
So 25% of the population has to be fully shielded for maybe six to twelve months ? You also have to make sure that anyone caring for the subset of that 25% isn't going to infect them. So, at a minimum, that's medics and those working in social care who, though not totally shielding, will have to isolate themselves from the rest of society to some extent.
Then you have the issue that the 75% of the population not shielding won't all get infected, because that's what natural herd immunity means and there'll always be a bit of transmission going on.
So, when you release the shielded population, the virus will then have new hosts to infect, which it would. The only way to stop that happening is to do what New Zealand did and persue an elimination strategy, so that's closed borders for the foreseeable.
As plans go, although at first glance it's an appealing one, once you scratch the surface you start to see the large holes in it, holes which no-one who argues for that plan has put forward a coherent plan for filling.
If you have a coherent plan, then spit it out mate, because you could save our economy from a pretty dire situation with such a plan.