I am assuming that for devices directly connected by zigbee/zwave to HE the following applies:-
Device Name is the abbreviated unique device name
For me
T01 is Thermostat 01
TRV01 is TRV 01
M01 - Motion Sensor 01
Temp01 - Temperature Sensor 01
Multi01 - Multi Sensor 01
and
Device Label is
Room - Function
in mine
Lounge - Temp Sensor
Bath - TRV
Etc etc
Does that
A. Sound reasonable in HE
B. Retentive 
C. Both
This is very similar to what I do, though I haven’t been abbreviating.
<Device type/function>
I have seen some also add emoji icons - that looks really nice. Not sure how it sorts though...
In my opinion, it makes the devices much easier to find which is especially useful when using them in apps.
Why use such short names and abbreviations when there is absolutely no need to?
But whatever, I'd say there is no "best" except whatever works for you.
If you plan to use a voice assistant such as Alexa, HomeKit, Google, etc you may want to name devices you plan to control with those systems something you can easily remember and those devices can easily interpret.
I do prefix some devices with Z, Y, and other letters at the end of the alphabet for ones I never plan to control directly or that often.
I think a system, and one that you stick to, is a very good start, it makes it much easier to find things when you know the pattern you would have used to name something.
I don't tend to use both the name and the label, I typically just fill out the name and leave the label blank, but would be interested to hear if there is a use for having these two...?
I normally include the room at the start and then some kind of name for the device itself, but can't claim to have been diligent enough to follow this now that I look though my device list.
Function-Room, like
light-kitchen.
This way when we search we always know what device type, and they are sorted by room
light-kitchen
light-MBR
light-porch
This is what's worked well for me
you could do vice-versa. Google even knows when i say "Kitchen light". Sometimes if I say turn on light-kitchen, it responds before I finish and turns on all lights!
I use the Label to name the device as I want to see it in logs, RM, etc. I then use the Name to give me device details/descriptions: Manufacturer, Model Number, etc. (I'm debating on whether to add last battery change date in there as well. And, yes, somebody will accuse me of being too detailed. 
)
Fantastic - that you guys.
Answers to some of the questions raised.
@morningz I am using the Device Name as the Unique name for the device as the * on the Device sheet states the Device Name and Device Network Id have to be unique.
And abbreviated because the Device Name is what I print on a Dymo machine to stick on the device.
(particularly as I am renting and that means I will be moving and sorting it out without labels would be tough )
@ritchierich Thank you for raising this point - EXACTLY why I raised the question - to trap unforeseen issues with taxonomy. You raise a good point I have not got into with HE yet - Which attribute does Echo use the Device Name or Device Label use - Because I am Ok if it is Device Label - because realistically I would expect to create a virtual Device and expose that to Echo, 99% of the time we only use scenes or 'Alexa turn on the printer' 'Alexa Goodnight' 'Alexa Turn on Freshen Up' 'Alexa Turn on Away Mode' etc - very rare to expose a device more an action.
@Rxich an interesting point here. This is a draw back HE, other HAs I have played with have allowed users to group devices together, In my case it was generally by room/zone but I did gather specific functionalities together as a group such as Heating (all things relating to heating) another one for Audio Visual devices, Speakers, Power usage Monitors etc.
Don't have that optioning HE so it makes the naming convention all the more important.
As long as I am consistent in naming - I can use the Device Search to create an ad hoc Group of say TRVs or Controllers so I think I am ok.
@JB_TX I like the way you do it but tough to print out a label that will fit on some of the devices
Something I really like in HE is quite a few of the Devices have a Date of last battery change built into them - really helpful.
My current system:
icon Room [location] object
where
icon = single character for "class" of object
"╚" for Lutron, "~" for Zigbee,
"⌐" for locks, "■" for virtuals, etc.
Room = room/area containing device, always capitalized
Location = *optional* location of device, if necessary
i.e., "N" or "S" for identical ceiling lights
object = description of object
"light", "chandelier", "switch", "sensor", etc.
Examples:
╚ Dining Room chandelier
~ Basement E wall sensor
⌐ Side Door lock
■ VACATION
The key is consistency - looking at an object in the field, you should be able to know exactly what the object will be named on your hub(s).
I'll echo the consistency theme, but also add that if you're using a voice assistant the command you use to interact doesn't necessarily need to contain the same name; i.e. if I have a device called Front Room Zigbee Outlet 46 Left the Alexa command could be as simple as Turn on the Blue Lamp - the routine needs the exact reference, but the UI doesn't always (there are ways of overriding this for dashboards using CSS too)
Label if specified otherwise name. Once you link them to Alexa, HomeKit, etc you should be able to rename them there too if that helps. But at least with Alexa you don’t have to say exact name as she usually figures it out.
This is a very important lesson for the user experience of the entire household, or so I have heard.
Just one hint that makes things in HE easier. I’m using Emoji’s at the beginning of device name to identify the device type.
For Rule I use the same principal, based on what is the Rule about.
By this approach, you have device types group together and you can easily find them. As a side benefit, you have those emoji visible in dashboard tiles. Also notifications can be enhanced by emoji to see what is happening without need of reading.
Do the emojis interfere with voice command with a Google Home?
I have learned that when creating scenes and groups, I can put a [S] or [G] in front of it and still use voice command with Google Home. It is able to understand to ignore the bracket letter. So I'm also adding LED lighting scenes, but start those with an [L]. This way from a list sorting, they are together.
So wondering, if Google ignores the bracket letters, I guess it also ignores the emoji?
I guess I need to look up how I'd add emojis to my names now...
No, I can still controll it via google voice command. However google reads emoji back in response "OK, turning lightbulb Martin light ON".
@igor.lehotan thank you. When I use the [L] and [G] it does not read those back. If I have a group of [G] Kitchen lights, and tell google to turn on the kitchen lights, it says, "turning on kitchen lights".
For the emoji, I wonder if brackets would make Google Home ignore saying it if it had a bracket around it?
I might play around by expanding the [G] into a word [Group] and see if that changes the behavior, but I like that Google "ignores" what is in the bracket.
Subtopic - The consequences of changing those names & labels later
Can one expect FULL propagation of all name/label changes through all apps & rules after being changed/updated later after the initial configuration.
I just went through some two step with one Sonos device shared on the same LAN by two hubs where I went in and changed the respective names/labels on the respective hubs to make them MORE specific to the hub doing the talking through the Sonos ( I HAD named everything exactly the same initially, then decided I might not want ANY naming/labeling the same between hubs on a shared LAN as and when I start to mesh devices or use Hub Connect etc).
I had something come up that made me wonder if changing the names/labels later on was detrimental to any Apps & Rules set up with the initial names. Not worth elaborating here because I didn't take notes on what all I had done that might have been material to the issue.
But in a generic sense, I thought I'd raise this question....IS THERE A DANGER in changing this stuff way later after it's been referred to in rules and other apps????
Rules and apps typically reference a device via its database ID not it’s name. This is the ID you see in the URL such as device/edit/258. Integrations typically use the device network ID set within the device. With Sonos specifically you should find the device network ID set to the hex value of its IP address.
I have renamed devices many times and even on meshed devices and haven’t run into any issues.
Thanks for the clarification.
So, does the often discussed "database corruption" exhibit the scrambling of some, or all, of those IDs ...or even an "old name" being presented as a part of the symptom?
No because the device ID is basically the row ID in the database - think an excel spreadsheet with the row numbers on the left side. That won't change but database corruption could affect a row causing issues but restoring the database will replace it back.
Download the Hubitat app
