Divide & conquer COVID testing strategy

[fikisipi]

What's your opinion on group testing? Here's a math analysis on it:

https://members.loria.fr/ADeleforge/the-maths-of-pool-testing-mixing-samples-to-speed-up-covid-19-detection/

Infection rate ► Initial mini-pool size ▼	0.5%	1%	2%	3%	5%	8%	10%	20%	50%
2	0.51	0.51	0.53	0.54	0.57	0.62	0.64	0.78	1.12
4	0.26	0.28	0.31	0.34	0.39	0.48	0.53	0.77	1.30
8	0.15	0.17	0.21	0.25	0.33	0.45	0.52	0.82	1.41
16	0.09	0.12	0.18	0.23	0.33	0.46	0.54	0.87	1.47
32	0.07	0.10	0.17	0.23	0.34	0.48	0.57	0.90	1.51

My concerns are:

• RT-PCR has 60-80% sensitivity which would ruin the predictive value of a branch (as you go up the tree the test outcome affects the final result bits exponentially).
• Is the lack of sensitivity due to (individual vs mixed sensitivity):
  a. virus RNA doesn't stay in everyone's throat
  b. failure in extraction/probing
• Can you reliably keep split one patient's sample into N parts where N is the testing tree height? (since max SampleCnt(Patient_i) = N)