Since dynamic markings do not have an absolute meaning (historically) it's hard to give a simple answer to your question.
Your use of the term absolute is "relative"! Relative to the audience which is not an absolute measure but an opinion (group consensus?). Different people in the audience may perceive the music too loud while other too soft.
Having worked in orchestras I can say that the conductor would often go into the audience seats and come back to adjust our volumes. When an entire section of horns would all play mezzo-piano (in their own mind) the sum of the sound would produce a mezzo-forte or mezzo at the conductor's seat and we would be instructed to play piano or softer to create mp. So, clearly the intent is to produce the effect for an audience. In that type of setting the audience is the most important participant. I would guess that composers are creating form a listener's perspective (i.e. "I want the audience to hear it like this..."). When playing solo the situation is different from a personal physical perspective but the intent is likely the same (to produce a certain sound for the listener).
When I say that dynamics are not absolute what I mean is that piano is simply softer than mezzo and forte, but that does not translate to a fixed decibel level from the point of view of scientific measurement. So if the loudest section of a piece is forte a player may play as loud as possible, but if the loudest section is ff they would need to play the f sections softer than their personal max volume. It is all relative which gives the performer some leeway in interpretation. Again, in an orchestra setting you might hear the conductor say "don't push too hard on the forte section, we have fff forzando at the end".
I think the intent is to produce a sound that is interpreted by the audience. To this end we are trained to try our best to control the instruments so that we can create a constant volume across frequency ranges. So if volume increases with frequency one might be instructed to soften up in the upper register of the instrument and not "overdrive" it. The contra bass is a particular issue as in the lowest register of the instrument it gets quite quiet.
So, your two examples actually feed the same goal, i.e. to produce a consistent volume at the listener's ear.