Meetings is an an interesting issue. Getting people together and focus on something is usually quit productive. We have a lot of those. War-rooms when we spend a day working on a specific data-set, journal clubs when we get together to examine a published paper, the ATLAS-days, methods-days when we spend a day going over our laboratory-methods and try to come up with ways to make them more efficient, database-days when we get together to update the ATLAS-database, writing-sessions when we focus two days on a specific paper, lab-meetings, etc, etc.
And on these meetings, we are focused on one thing, usually something later-stage-research, which is not data-production. No data is produced. Thus, a week with two ATLAS-days, a journal club and a lab-meeting for example. That would be an incredible week for adding competence to the group. But not much data produced that week. The machines are standing still. Less data to have meetings around next week. And after a few such meeting-weeks we can all go on vacation as there is no data anymore to work with. Some 20 years ago I briefly worked at a place where someone calculated that 51% of the time was spent on meetings, and only 49% on the actual data-production, what we were hired to do.
Some meetings are necessary, and (actually most) are really really good. But the best way of checking that we do not have too many of them is to confirm every meeting-idea I have with the researchers in the group. They want to produce good research, and usually when I air ideas about too many and too weird meetings, they will stop me and tell me enough is enough.