I feel like they're trying to make an insidious suggestion about the usage of these. IMO, there's likely a good reason - user experience.
At a hardware level, grabbing the microphone can take time. Even worse that timing is inconsistent across devices, workloads, etc. That leads to a bad experience when unmuting and needing to delay your commentary. The solution to this is to keep the microphone on, but mute at a software level. This way the mic is always hot and ready to relay audio as fast as the software can switch.
I'd be somewhat willing to bet continue to stream audio is also a quality assurance mechanism. Some networks will shape traffic according to load. A quick jump in bandwidth can introduce unexpected jitter and latency. By continuing to stream audio (but not necessarily process or re-transmit), video conferencing can better ensure an un-interupted experience.
----
With that being said, if you really care about privacy, consider getting a hardware mute microphone.
At a hardware level, grabbing the microphone can take time. Even worse that timing is inconsistent across devices, workloads, etc. That leads to a bad experience when unmuting and needing to delay your commentary. The solution to this is to keep the microphone on, but mute at a software level. This way the mic is always hot and ready to relay audio as fast as the software can switch.
I'd be somewhat willing to bet continue to stream audio is also a quality assurance mechanism. Some networks will shape traffic according to load. A quick jump in bandwidth can introduce unexpected jitter and latency. By continuing to stream audio (but not necessarily process or re-transmit), video conferencing can better ensure an un-interupted experience.
----
With that being said, if you really care about privacy, consider getting a hardware mute microphone.