Great discussions!
Reading comments on this site is like reading the take-home-message of an academic book, very useful and distilled information.
My two cents:
Laboratory blood tests are a treasure trove, I agree, I hope somewhere some expert modeller is able to sit together with a biologist/biochemist to develop a good prediction model from those data, BUT
at this point of the pandemic I believe we rather need a model for testing/sampling campaigns, while relying on precision of the available diagnostic tests which seem to be quite good already (provided RT-PCR is widely accessible)
I think this decision should be left for the ethics committee to decide, if waiver of consent is in the best interest of the public. The researcher SHOULD still submit an ethics application for approval to consider their case to waive consent. This process can be expedited if time is of essence.
Ethics committee is the last defence for potential participants to prevent exploitation, also serves as an eye to make sure the research proposal/methodology is ethically sound. A large study without ethics or waiver of ethics can be problematic - see the recent controversial hydroxycholoroquine study: https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(20)31180-6/fulltext#seccestitle80