Do the stats measure parallel productivity?
Let’s say a well-meaning participant gets invited to a meeting but his active participation is needed for only 5 minutes out of the 45 scheduled for the meeting, so for the other 40 minutes he either “arrived late” or “left early” or “be on mute with video off” and still remain productive working on other things.
I’d be very surprised if a company doing its measurements based only on what’s recorded in virtual meetings can get to these things that are not recorded by virtual meeting tools. This well-meaning participant would just contribute to the “arrive late” or “leave early” or “mute/video off” statistics.
Before you ask “why invite someone to a 45 minute meeting when he’s only needed for 5”, the practical answer is there are many things for which a 5-minute real-time interactive Q&A can resolve a lot more than back-and-forth group emails that take over 50 minutes to compose.