At a quick reading, here are some things that pop out: 1) In your 94 well example, you'd need to dilute each individual sample 1/94. This could cause problems.

At a quick reading, here are some things that pop out: 1) In your 94 well example, you'd need to dilute each individual sample 1/94. This could cause problems. The lower leveraged scenarios may be more useful due to lower dilution. 2) Similarly, very small scale liquid handling is relatively difficult to do. As you multiplex the number of sources into a well, the volume of each of those sources will decrease, increasing the difficulty and uncertain. Once again, a lower leveraged solution is likely viable. It's also worth identifying that the bottle neck in testing is moving towards the availability of sample swabs, and reverse-transcription reagent (these happen up stream of PCR). Given those constraints, the emphasis should be on making all other steps in the detection chain as reliable (so simple) as possible to make the best of the bottle necked resources. That said, it's a very clever idea.