It would help if you could clarify what you mean by "increased costs" (professional fees for geotechnical engineering, structural engineering, construction costs, etc?).
The answer(s) depend on the project location, soil/rock profile and the Seismic Use Group / Seismic Design Categories. We have found that essential facilities and structures which represent substantial hazard to human life can easily justify the relatively small costs to accurately determine the Seismic Site Class. Where such facilities are located in Site Class C, D, E or F areas, we have found on several occasions that field shear wave velocity measurements have saved several hundred thousand dollars in structural detailing and construction costs. Larger facilities have seen even larger savings from better soils data. Accurate soil profiling, using CPT, was essential in screening whether it was likely that shear wave velocity testing would yield higher site classifications.
With regard to deep foundations, we typically assume for site class purposes that the soil/rock profile begins at the bottom of the pile cap. The simple explanation is that the deep foundations are typically slender and flexible compared to the free-field motion and thus will not usually significantly reduce the base shear felt at the pile cap (and subsequently transferred to the structure). There are obvious exceptions to this, and in some cases, a full soil-pile-structure interaction analysis is justified.
As geotechnical engineers, we should do our best to educate the owners/structural engineers that it is in their (and our) best interests to get the data necessary to adequately characterize the soil/rock profile for the intended purpose.