What is defined as "reference" level of a speaker? Is it an absolute volume (ie. 95dB) or a relative/subjective level?
Most people feel that a reference speaker is an immaculate, high-end speaker used by studios as a transparent monitor.
'Reference' is a very broad term used in different contexts in the AV world.
What 'Reference' in general means for an AV setup is that, it creates the ability for the listener to hear/see exactly what the creator(s) intended. In the broad sense, all of the interpretations can be categorized as one of two; Quality and Standard.
Reference Quality:
Something considered by an authority or by general consent as a basis of comparison, which is, regarding quality of recordings, electronics, speakers, rooms, etc. Here the “authority” is scientific measurement and ‘general consent’ can be found in the ‘Software, Movies & Music’ section of the forum.
But what does that mean in the real world?
Source: This varies depending on media and recording and is achieved more from anecdotal referrals rather than measured/ tangible quantities. It means, all else being equal, the recording helps understand if a setup is good or bad. Or, it will sound good on a good setup and bad on a bad one.
Electronics/ speaker: Equipment that does not change the signal when compared to the source. So for a speaker, AVR, pre-pro, amp, TV, PJ, etc. it means, they need to have a flat frequency response between the max and min range of the signal. On the audio side, oversimplifying it, if the creator(s) want you to hear 400Hz at 75dB, your setup plays it as such. On the video side, if the creator intended you to see brightness of 75% max for color 25% Red, 30% Blue and 45% Green, your display produces it as such.
Room: Now this one is the most difficult to create since it is a combination of the above and needs to account for room factors. On the audio side, it needs to have a minimum ambient noise level and not boost or cut frequencies due to overlap as sound waves bounce between the boundaries (walls). On the video side, the ambient light should not effect yours eyes in a manner that the perceived video characteristics change. Light control for a room is easy, make the walls dark and keep the blinds closed

.
Reference Standard:
A rule or principle that is used as a basis for judgment. Something that stays consistent irrespective of your measurement principle and tools.
But what does that mean in the real world?
This is easier to explain. It is like saying, "my Reference Standard for time is Seconds". Since, determination, measurement and communication of what we mean by a 'Second' in scientific terms must/will be identical.
On the audio side, Dolby, DD, DD-HD, THX, THX-Ultra, THX-Select, DTS, DTS-ES, DTS-MA, etc. are all 'Reference Standards' and on the video side, they are PAL, NTSE, RGB, etc. NOTE: What is not Reference Standard are the various "matrixed sound processing" options offered by AVRs like, "Music", "Game" or "Stadium" or "Club", etc. since each manufacturers interpretation and implementation of this nebulous sound will be different. Or for the same reason, on displays, "Dynamic", "Sports", etc. never correspond with a Reference Standard for displays.
What this means is, any creator(s) of media will have to adhere to specific rules when making recordings for a standard of their choice. But, it does not end there. Both Dolby and THX also have specifications for electronics, speakers and rooms. Meaning, if an amp is THX "approved" and sports their ubiquitous logo, it has met certain specifications for what it does with the incoming signal, the components used, the design and implementation of the circuits, etc.
-------------
Sidebar:
Let's say, I make the some milkshake and the world and agrees that it brings the most boys to the yard, so, is the best milkshake ever. So, we can say, my milkshake is Reference Quality and 'bringing boys to the yard' is the Reference Standard. You may not agree that my milkshake is the best, but since the whole world thinks so, I can say,
My milkshake brings all the boys to the yard,
And they're like
It's better than yours,
Damn right it's better than yours,
I can teach you,
But I have to charge
And that is exactly why all the mass market equipment is so expensive and it means jack isht if they are actually good or not.
-------------
But, in all this, how do you know that your setup is 'Reference Quality and Standard'?. This is done via measurements and tweaking settings. This process is termed 'Calibration'. How to calibrate for reference is a whole other discussion, but, there are several articles under AH University that guides a user through this.
To be continued... (need to drive home

)