That's excellent. You can also use some non-Hue bulbs on the Hue hub. The Hue bulbs are great but admittedly pricey. Not sure if the GE can or not.
A Zigbee bulb that has been powered off will most definitely not function as a repeater, until it is turned back on.
Maybe I could try cutting power to those bulbs by turning the switch off, then, and see if that changes things after I reboot the hub.
It takes a bit of time for a zigbee mesh to rebuild itself. How long depends a lot on your network. I'd try removing them for a few hours (or maybe overnight) at the least.
Absolutely. Just like in real estate, it's location, location, location. You could have a zigbee outlet behind a refrigerator that makes a pretty lousy repeater while the same device in an area where it is not blocked by metal could work great. You never quite know what's happening to those radio waves as they zip around your house.
The trouble is, some of these repeaters need to be where they are. They don't have to function as repeaters, I just need them to be smart plugs. But... if they are in bad locations for repeaters, will they bring down my Zigbee network? Because I don't know what to do if every Zigbee plug is a repeater. Some of those have to be in certain hard to reach places...
I think over time your hub will find better repeaters and that's what will show up in your neighbors list.
Can you actually unscrew them, so if they are accidentally turned on, they don't immediately rejoin the mesh?
Hopefully Hubitat hubs are smarter than Hubitat owners. I'm counting on that in my case...
Nope. I think your Link bulbs interfere with your zigbee 3.0 plugs.
It's a learning curve like anything else. I still learn something new every day, usually by reading what @aaiyar and @ogiewon post!
Yes, I just need to grab a ladder. I was just thinking may be initially.
If the bulbs are simply unpaired but left in place, would they still join the mesh and cause problems?
You need to reconsider that. Seriously!
No... if they're not paired they'll work just like... lightbulbs. Well, maybe flaky lightbulbs 
They should not. But I've accidentally paired Link bulbs in my neighbors house .....
To elaborate a bit on this:
Zigbee routing links are ranked to give preference to paths which are good in both directions, since the protocol of sending a routed message always involves some bidirectional communication. So the hub needs to hear a good signal from its next routing hop, and likewise that routing device needs to receive a strong signal from the hub.
The LQI figure shown in the table tells part of the story: it's a figure of merit for how well the hub 'hears' a given device based on observed signal strength and/or link error rate; the spec allows each vendor to derive their own metric. The routing algorithm maps this LQI number into a single digit InCost value (lower non-zero numbers are better; if they are zero that means the cost info is stale or missing). But it tells only half the story; it gives no clue as to how well the device at the other end of the link receives transmissions from the hub. The hub doesn't want to route messages through a device that has a low probability of receiving them.
So to let the hub know how well the outboard device hears it, that device periodically (4x a minute, regardless of whether there are messages to be routed) transmits a figure of merit derived from its own computed LQI in a 'link status' message. The hub shows this as the outCost value in the table.
Since the number of routers that a hub can keep track of is limited to 16, the hub can choose the best (based on low inCost and outCost figures) to maintain in its neighbor table. The neighbor table can change based on the current state of the mesh; this is part of what makes Zigbee self healing; each routing device also maintains its own neighbor table in a similar manner.
That's a great explanation... thanks!
So I unpaired those four GE Zigbee bulbs, physically removed and replaced them with dumb bulbs. They are completely out of the picture now.
I can't say for certain that there was an improvement in my Zigbee mesh. The Zigbee blinds drivers that I use seem more responsive, but I can't tell for sure since their performance has always been hit or miss.
In the interim, I upgraded Hubitat to the latest version.
I reran the output of the zigbee Child/Route Info. Here is what I got:
Parent child parameters
EzspGetParentChildParametersResponse [childCount=0, parentEui64=0000000000000000, parentNodeId=65535]
Child Data
Neighbor Table Entry
[Tuya Zigbee Repeater, 3FDA], LQI:248, age:4, inCost:3, outCost:1
[Master Bedroom Diffuser, 78E4], LQI:236, age:7, inCost:5, outCost:0
[Holiday Lites Dining Room, 9A89], LQI:244, age:4, inCost:5, outCost:7
[Master Bathroom Diffuser, AAC4], LQI:82, age:5, inCost:7, outCost:7
[Game Room Table Light, C73B], LQI:251, age:3, inCost:3, outCost:1
[Holiday Lites Temple, CA89], LQI:203, age:7, inCost:5, outCost:0
Route Table Entry
status:Unused
status:Active, age:64, routeRecordState:0, concentratorType:None, [Tuya Zigbee Repeater, 3FDA] via [Game Room Table Light, C73B]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Garage Sensor, 4A46] via [Game Room Table Light, C73B]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Master Bedroom Diffuser, 78E4] via [Tuya Zigbee Repeater, 3FDA]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Holiday Lites Temple, CA89] via [Tuya Zigbee Repeater, 3FDA]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Utility Room Sensor, 29EB] via [Game Room Table Light, C73B]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Powder Room Sensor, 7045] via [Tuya Zigbee Repeater, 3FDA]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Holiday Lites Dining Room, 9A89] via [Tuya Zigbee Repeater, 3FDA]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Master Shower Sensor, 715A] via [Tuya Zigbee Repeater, 3FDA]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Master Bathroom Diffuser, AAC4] via [Tuya Zigbee Repeater, 3FDA]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Kids Bedroom Sensor, C33B] via [Tuya Zigbee Repeater, 3FDA]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Master Closet Sensor, 0323] via [Tuya Zigbee Repeater, 3FDA]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Kitchenette Sensor, C896] via [Tuya Zigbee Repeater, 3FDA]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Master Bathroom Sensor, 7FBB] via [Tuya Zigbee Repeater, 3FDA]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Game Room Table Light, C73B] via [Tuya Zigbee Repeater, 3FDA]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Kids Bathroom Sensor, 70D2] via [Tuya Zigbee Repeater, 3FDA]
Addendum to above: I added another Luntek Zigbee Smart Plug, calling it "Study Room Zigbee Smart Plug." I reran the above and got the following for that one:
[Study Room Zigbee Plug, F430], LQI:212, age:7, inCost:5, outCost:0
I assume this would be considered a bad location for this repeater, correct?
Let the mesh settle for a little while. Had you turned off the Hubitat for 20 minutes?
Yes, I turned it off for about 30 minutes in fact before I started testing again. On a positive note, I didn't have to repair anything.
I didn't turn it off again after installing the newest Plug in the study room. I guess I should probably do that, too before concluding that this location is no good.
Download the Hubitat app
