The author is thinking that reversing a deleted zone mutation is hard, but it could be that a simple mutation of the weak strain make it deadly, so my question is: can you guarantee that a new mutation is not going to make this strain deadly?, also the more people with the new strain of the virus, the more the bad mutation could happen.

So I suggest looking for a strain for which we have a mean of killing the virus, that is a weak virus for which we have a general vaccine for expected mutations, if that is possible.