A misunderstanding is that this system results in instantaneous contact notification and isolation (which is what one of the original papers on digital contact tracing efficacy assumes).

The reality is the entire 'human in the loop' contact tracing process is still manual. Health staff working in state contact tracing teams still need to call and interview every contact to determine whether they could be considered a close contact from an epidemiological perspective.

Also consider that close contacts include anyone you've spent more than 2 hours in an enclosed room with.

So that immediately limits the utility of any bluetooth proximity app: you still need an entirely manual process for home, work, social gatherings - any situation where you are in an enclosed room.

We've taken a deeper look at the assumptions and expert opinion, both from individuals and institutions, which appear to discount this kind of system from providing significant value.

Even the product lead of Singapore's TraceTogether app, as a natural advocate for this kind of initiative, admits the technology is oversold and is only an additional tool (due to the significant potential for false positives / negatives):

https://blog.gds-gov.tech/automated-contact-tracing-is-not-a...

Thanks for the insight.

Why wouldn't they just auto-alert anyone who came in adequate proximity to get tested though? Surely there is some benefit to letting people know they need to be extra careful now and get tested ASAP.

All the manual human work can continue as normal.

If someone is determined as a close contact they have to legally self-isolate for 14 days regardless of symptoms or test results.

The policy there probably reflects the understanding that people in the early stages of infection won't test positive or exhibit symptoms.