I admit - I am not an active user of Zoe.
I understand the premise - a subset of the population (a minority) report their symptoms regularly which provides enough data to make wider predictions and assessments.
When I looked at my area on Zoe there were so few people reporting compared to the overall population that it would be daft and unnecessary to use such a small dataset to make any prediction - the huge number turning up to be tested was (IMO) a far more accurate way to estimate virus patterns locally.
Zoe is an interesting concept but I am not sure it is as relevant as it was last Spring - why rely on something that is an estimate based on a small minority sample of the population, when the entire population has access to testing when they deem that they have symptoms (a far, far bigger sample than those reporting on Zoe).
I suspect Zoe was hugely useful earlier in the pandemic when govt data was patchy and testing capacity was very, very low. People with symptoms couldn't get tests then. So we couldn't get a clear picture of what was happening anywhere with the available data. So an app like Zoe was invaluable.
But that has changed.
Now the testing capacity in the UK is phenomenal and results are available within a few days. So I am not sure what Zoe can meaningfully "predict" accurately that the govt published data cannot.
And my guess is, that a lot of the time, it's quite a bit off - because it is an estimate based on a minority of the population electing to report, of which an even smaller minority have symptoms.
Just my tuppence..... 