I'd like for high risk biological research to be done under the same security arrangements as nuclear weapons research. Covid and influenza have together killed several orders of magnitude more humans in the last 150 years than nuclear weapons/accidents have. Thus GoF research presents a potential human cost much higher than nuclear weapons, and should be treated as such.
If that makes it "too expensive and slow" then so be it. Perhaps there are some things humans should not be doing.
Well, historically, you know, nuclear weapons research involved a lot of cavalier scientists doing dangerous things with limited knowledge, and of course in the early days, there was plenty of above-ground range testing; is that something you want included as well?
If my information is correct, I believe that nearly all nuclear testing these days is 100% simulation on supercomputers, which is as safe as safe can be, as long as you can trust supercomputer simulations to render the correct answer to those trusting researchers.
If that makes it "too expensive and slow" then so be it. Perhaps there are some things humans should not be doing.