Why are my Cubase meters giving inconsistent readings when playing back the same snare hit on loop?
-7.0 to to -1.3 is a big jump 
Is this something with my setup? instability?
Depends on what you are using.
Is this a wave file loop?
Using midi to trigger instrument?
Using plugins?
Using something like beat designer with a variation algo?
Variation algo on the plugin instrument?
Wich meters are you talking about, track, track/output, track in/track out, pre/post fader, pre/post sends…
As you can see, there are many things that can caus different readings at same samples.
You have to be a bit more spesific. 
I would hope it would be obvious to anyone, that this would be tested best with a looped audio .wav file.
It’s been reported before:
I’ve also noticed that my Stereo Out, is 2db louder, than the single mono source being fed into it… Shouldn’t it be the same?
Dont take this up the wrong way, but the tech level and up to speed, is not constant between users.
So if you posted the links on the first post, it would be much more easy to give opinion on the matter 
I have never notice this issue running the 32 bit engine. I havent bothered with the 64 bit. Dont have enough Ram to run that engine.
Major drawback with channel meter in cubase, is that it is post fader. Wich make it a bit so, so… when you start using the faders and do adjustments to the channel pre fader. But stil, the level should be constant at fixed level. And yes a mono output should read the same on the master at master fader set to 0.
But if you have adjusted the master fader, so it no longer is at 0. You will get different reading.
I’m pretty sure everything is set to 0, and there is still a 2dB difference
Pan Law ?
I guess, I always though Equal Power was… Equal Power… I guess I should read up on what Equal Power actually is now.
This doesn’t explain the OP problem though of inconsistent meters in 64bit mix mode.
I’m actually finding some other inconsistencies with VUs, I’m creating a bit of a sample library and doing pretty focused detailed editing, level matching etc… And it seems like all the VUs can intermittently vary by 0.2db to 0.3db - I pretty much have resorted to having to double check the level 2 or 3 times by playing back the sample 2 or 3 times.
This is in 32bit mode, so it seems the problem becomes worse in 64bit mode.
Amature stuff from Steinberg imo.
Can someone else try duplicating this?
To test efficiently, assign a hotkey to ‘Clear All Meters’
Set meters to input
Take a snare sample and make note of its db value (or adjust it to perfectly be -6db) and repeat it once every bar (to give yourself space to clear the meters) 8 copies. Then duplicate this track three times, so that there are 3 tracks of the exact same snare sound, played 8 times in unison with each other.
Open your mixer so you can see three meters at once, press play, after each snare hit, clear the meters and see if there is any deviation away from -6db.