Hello @billg58
This is, unfortunately, complicated.
WiFi client devices decide which access point to connect to, which band to connect to (e.g., 2.4GHz or 5GHz) if it supports multiple bands, and when to switch between access points or between bands.
Google WiFi and Nest WiFi (including Pro) do provide clients with some "extra" information in the form of 802.11k and 802.11v that WiFi clients can use to make much better decisions about when to switch and what to switch to.
However, these specifications are optional, and many WiFi client devices do not implement them. Many phones do (but not all), and some laptops do (but far from all). But, most fixed devices (e.g., smart TVs, streaming sticks/boxes, gaming consoles, smart speakers/displays, smart plugs/bulbs, cameras, doorbells, and other IoT devices) do not support these specifications.
Devices that do not support these may still switch eventually. I've seen some Amazon Echo devices switch after about 24 hours, for example (but consider that anecdotal). But most will just stay "stuck" to what worked when they first started up (or worse – see below).
Things get way more complicated at this point, though. There are so many different ways people have implemented clients. Some may do a selection on startup, with a full scan of available access points and bands before selecting the "best" option. These devices work pretty well, and they can get connected to an optimal access point just by restarting the device itself.
However, some devices may just start listening on startup and connect immediately to the first beacon they hear. These devices effectively connect to a random access point. Restarting them repeatedly may get them connected to a more optimal access point.
But, the worst devices do a selection only once – when they are configured. They remember it (channel, BSSID) and as long as that AP still works, they just connect to it every time they start up, not even looking for another option. If you are dealing with devices like this, they may need to be reset and reconfigured to get connected to a better access point.
Ok, so here's where it gets worse. When you restart the WiFi network, the primary/router unit comes up first, followed by the secondary/point units. So, during a network restart, all the devices are looking for a network to become available. They will typically connect immediately to the first access point that becomes available. Since that is the primary/router unit, fixed devices very often end up connected to it. Due to everything I described above, those fixed devices end up stuck to that distant access point even after the better option comes up, even when that happens just a fraction of a second later.
Whew. Sorry for all the details, but like said, it's complicated. My advice would be to restart all of your fixed client devices after a network restart. Hopefully that gets most of them connected to a closer access point. You may have to do that every time you restart the WiFi network. So, try to avoid doing that. Most of the above discussion applies to most recent WiFi networks that support multiple access points, by the way. This stuff is so complicated I am sometimes surprised it works as well as it does. But, understanding a bit of the underlying details can help explain some of the odd behavior you may see. I hope some of this explanation is useful!
