I agree there can be many confounding factors. To clarify your point: this leads you to the conclusion that the asymptomatic proportion is likely under-estimat

I agree there can be many confounding factors. To clarify your point: this leads you to the conclusion that the asymptomatic proportion is likely under-estimated? And conversely, the symptomatic proportion is likely over-estimated?