Sorry if it's a stupid question, but hypothetically, would we be able to engineer rather than find a harmless strain of coronavirus with the spike protein of SARS-CoV-2?

If I understand correctly, antigens identify the virus by its S proteins. Would we be able to use the same methods <a href='https://www.nature.com/articles/nm.3985'>as in this paper</a> to replace the S protein in a harmless cold-like coronavirus with the S protein of SARS-CoV-2, and would then the resulting immunity defend against SARS-CoV-2?

This is not a stupid question. Yes we could engineer such a strain, the problem would be testing it given we would not be able to predict how dangerous it was.

Having identified a harmless strain, and presuming regulatory agencies are too slow to be useful, then what?

Send people infected with the harmless version around, after lockdown ends in a month or two, to crowded places?

Would greater virulence be a desirable quality in our reduced strain?

Well if the regulatory agencies decided that they weren’t going to act then I suspect there would be a grass roots level spread of the strain anyway. This would be the worst way to use such a strain, but I don’t think it would be possible to stop this happening.

I don’t think we will get to that point. If we collect good data from the natural spread of any attenuated strain then I think the regulatory authorities will allow its use on the basis of this data.