New to the activity regarding mastering. I have noted different advice given by folks in this forum regarding the proper output levels for mastering. Some have said -3db, others -6, and some as high as -0.3. Why the difference?
AND, surely the people who master the things can turn the volume down to where they need it to be so why the stickiness regarding the output level as long as it’s not clipping or distorting (running up to 0db or higher)?
It all depends how your music will be mastered.
You should leave some headroom if
a. Your music will be mastered in analog domain
b. Your music will be mastered in ancient fixed-point-engine DAW
c. Your track need to be converted into different sampling frequency before mastering
But if mastered digitally with floating-point DAW, there’s no need for headroom. Levels can be lowered in mastering process as required without any degradation of audio quality.
Why the difference?
In most cases: “because famous mastering engineer N.N. said it shoud be -XdB” or “I can hear the difference between -YdB and 0dB”. It’'s also about old habits remaining while there’s no any reason for them anymore.
Of course your should ask your mastering engineer. He/she may feel more confortable with souce material peaking at level he/she is used to.
Best bet would be to ask the Mastering engineer you plan to use what he/she prefers. 
Thanks for the insights. As I suspected, it’s more about what a particular Mastering Engineer wants.