This was posted on Canons repair forum. Not mine, but just reposting.
It's interesting but without analyzing the actual radio situation in that moment it's one empirical evidence. I routinely use Canon wireless in my home studio, where there is also a WiFi AP, without any issue. But I'm in a detached house and beyond my own WiFi other signals are quite weak. My APs are set to regulate the transmission power as needed, also, and that might help to avoid hampering the Canon signal. Also, wireless flash should not disrupt WiFi transmissions.
A building with a lot of APs filling each and every channel, blasting at full power, the interference can be far worse. Still, being the 2.4GHz band an unlicensed one, interference must be expected, and any design should account for them. Wireless flash has the issue it has to be a real-time signal, with strict tolerances, unlike WiFi which can stand far higher latency. Maybe today using a different band is not a bad idea, since most devices moved to WiFi, although many now uses the 5 GHz band.
Anyway, we don't know if it is a problem at the hardware level - the hardware enter a state when it loses the link and can't re-establish it until power cycled, or it's a software problem, the hardware still works correctly but the software is in a state when it cannot resume operations. The latter might be far simpler to fix than the former, although without software update capabilities Canon units can't be fixed that way.
I hope Canon is working on new, more powerful units and they might need to exchange more data - requiring a protocol overhaul.