I think the point about a vaccine effect is that (assuming there is one, and I don't see any reason why there wouldn't be)
a) it would show up in every single data series eventually, just at different rates and different times
b) it would show up more and more as time goes on

So, fine (in fact probably right and healthy) to be cautious and sceptical of e.g. @d_spiegel or @ChrisGiles_'s graphs today (and various others), but check back in every few days and see how they look then.

For now, I personally think there's a faint rumble of horses' hooves if you lay your ear to the ground, but it may not be long before nobody can ignore the arrival of the cavalry,

I think it was easier to call the start of deaths going up in September https://twitter.com/BristOliver/status/1306473804361678848 even when it wouldn't have passed any formal test of statistical significance, because we knew exactly when to expect it. But anyway, like I say, keep checking in.