measuring rests (for three)

I'm trying to work out how to structure the rhythmic shape of the piece (for three).
There are going to be a lot of rests, and I'm not sure how to measure them.
I have basically three ideas:

1. To use standard rhythmic notation;
2. To use minutes and seconds to mark the rests (but probably use standard rhythmic notation while they're playing);
3. To use breaths as a durational method for rests, and for the playing durations.

Idea 3. is very attractive to me, but I don't think it's for this piece. I'm thinking I want to use this in another piece. Maybe more on that later.

I think that each of these three methods of notating rests will create a marked difference for both performers and for audience. I would like to see some research on this (or I suppose I could actually research this myself I suppose). I've written pieces that use the first two methods of measuring rests and performed pieces that use all three, and I certainly feel that the experience is different.
Counting the rests creates (for me as a performer) a sense of very measured time.
Counting minutes and seconds normally involves watching a stopwatch - I would say that 'musical time' freezes and this correspondingly creates more of a sense of stasis for the audience as well.
Having the rests measured by somatic measurements creates a more intimate experience (at least as a performer - I haven't knowingly experienced this as an audience member) and I believe that this experience will be carried through to the audience.

But I don't want to do this for this piece.

Further refinement to the structure of the piece
There are three 'states' of operation:

1. Long measured silences frame statements of the three pitches (with various tunings)
2. Very few silences as the three instruments make up a 'tapestry' of sound from the three pitches
3. Long measured sustained single tones (staggered between the three instruments) frame statements of the three pitches

Rather than these being three different sections of the piece, I'm going to construct the whole piece in two parts. For the first part, I will structure rhythms it in two ways, using both methods 1. and 2. The finished version will progressively move from method 1. to method 2. The second half will use methods 2. and 3. in a corresponding way. This way, there are no clearly defined sections, but each moment of the piece should lead smoothly to the next.

In terms of pitch organisation, I'm hoping to have initial sketch material of the first section ready sometime tomorrow. I think that it's going to stay pretty restricted for the majority of the first section, and then start to move up and around in the second section before stabilising close to the end (although the whole range of the piece is going to be strictly within an octave - when you're exploring an expanded set of proportional tuning ratios, you don't have to gild the lily - which also serves the indeterminate instrumentation).