I interpret the IBC to indicate that foundation bearing elevations must be within 10 feet of rock (Vs >2,500 fps) to meet Site Class B requirements.

My question is the following:
Given a single story building relatively close to variable depth rock as defined above.

Can we lower all foundations to within 10-ft of site class b rock and meet the IBC code requirement/intent?  At what point would this technique not meet the code requirement/intent.......(1-ft, 4-ft, 15-ft)?

Are there exceptions to the code (e.g. can a small portion of the building not have foundations bearing within 10-ft; what if 5% of the building had foundations bearing within 12-ft of rock?

i would make that judgement based on the soil you're dealing with. i have lowered footings before to bear on rock and in spots it was probably 12-14 feet to the ground surface. however, it was high consistency soil with a foot or two of partially weathered rock.

the idea is to have a generalized profile satisfy those requirements but i would lean heavily to making sure most everywhere satisfies the requirements. so if you have one corner that is 14 feet (with "good" dirt) and the rest is 5 feet to rock, i'd lower the footings and roll with it.

i assume you actually measured the shear wave velocity at the site since that's the appropriate way to achieve site class b (and is required by code for site class b or a).

I am not familiar with IBC.   However, in most of the codes that I have been dealing with, the site seismic classification is up to the elevation of the lowest building slab.  Which means, lowering the footing is not going to change the condition much.  For example, in OBC, even if you put all of the footing over bedrock, if the overburden soil is soft or very soft clay, you are still looking at site class E or F.   

geoman, are the codes you are delaing with clear about the lowest building slab?

That may be sheer curiosity but some codes, like the Italian one, pose the reference level at the bottom of the foundation structure.

The common interpretation of Ontario Building Code (OBC) is that the reference is not lower base of the foundation and in most cases is the lower level of building (AKA top of foundation).  For example, if you have a site with thick soft clay and use pile foundations driven to the bedrock for support of the building, you can NOT consider the site as Class B.  Similarly, if the soft clay is medium thickness (say 3 meter) and you decide to use deep footing placed on the bedrock for the support of the building, it is still NOT a Class B.  

I believe the idea is that the pile and/or foundation wall (of the footing placed over bedrock) are not going to move completely independent of the clay mass surrounding the pile/foundation wall when earthquake hits.  There would be interaction and therefore, the soft clay mass will have some effect on the foundation movement. It would not be similar to a case when the footing is placed over bedrock with thin stiff overburden soil (hence not Class B).

On the other hand, as I mentioned this interpretation applies to most cases.   For example, in case of a building with multi-level basement and perimeter retaining walls being part of the structure, it is not completely correct to consider the lower level of the building.  In my opinion and by the same logic noted above, the perimeter retaining wall that are in contact with the structure and extend all the way to the ground surface would have some effect to the movement of the building and therefore, considering the lower level of the building for site classification may be incorrect (particularly, if the soil between the ground surface and foundation level is much weaker than the soil below foundation).  In this particular case, considering the ground surface may also be conservative.  In a case like this, it is better to conduct a detail site specific assessment to come up with the parameters (It may save some money). If the owner does not want to spend the cost of detail analyses, they you have no choice unless to assume it from ground surface.
 

That's pretty clear geoman,
in the Italian code there is a distinction between 'shallow foundations' (reference depth is base of foundations) and deep foundations (reference is top of piles or wall), whereas the intermediate cases are not contemplated.

When a shallow foundation becomes a deep one is also a matter of technical judgement, as a whole I agree with your observation.