A couple of IMPORTANT questions first:
Is this a “route traverse” or a “loop traverse?”
What is the total [topographic] RELIEF involved?
What is your distortion budget?
I should ask (but won't...because I don't want to start WWIII this morning) WHY you are using SPC (or UTM) to start with, IF you want to be “on the ground” when you are done! Besides which, I probably already know the answer to that one.
The “distortion budget” can be defined as simply the ACCEPTABLE magnitude of distortion within the final adjusted solution. Basically, the relationship of the final grid distances (assuming perfection thereof) and the actual ground distances. Unless you are working in a very flat terrain, this distortion can be anything from a few ppm to well over 100ppm.
If you have [say] 400 feet of vertical relief in your project, and you use the average height for your developed surface, then you can see as much as ~10ppm grid to ground distortion within your traverse (even AFTER it is adjusted). This would be acceptable for MOST practical applications. Once you get much past that threshold, it becomes noticeable to most careful surveyors, although I have found that even 30ppm falls within the warm-n-fuzzy of many folks, and 30ppm is acceptable for most construction projects too.
Any georeferenced coordinate projection will “tie to site relative to other sites in the vicinity,” so that's pretty much a matter of personal preference.
The traverse WILL have to be run in “3D” in any case. To do otherwise precludes the proper computation of vertical convergency, and will degrade the final results based on the amount of total vertical relief in the traverse. More relief, more degradation.
Once you have DEFINED a georeferenced Coordinate System for your Traverse, Star*Net will do all of the heavy lifting for you, including “moving” (expanding/contracting) each horizontal distance to the developed surface of your projection. Star*Net supports just about any formal mapping projection (I think), and can certainly handle any SPC, UTM, or LDP. I haven't used Star*Net in many years now, and I never used it on MODIFIED SPC, so I don't know how well that technique is supported.
IMO a Low Distortion Projection (LDP) specifically tailored to your project would be both the most accurate solution, and the path of least resistance, but that's just my personal preference. OPUS has already generated and returned all of the fancy data that you would need to generate the projection parameters (constants), and any decent coordinate transformation software (like Blue Marble's Geographic Calculator) will transform the LDP back and forth to any other georeferenced coordinate system (SPC/UTM/LDP) in the blink of an eye.
It is NOT my intention to lobby for an LDP here, and if you would like to post (or Email to LDOGEO at AOL dot COM) the OPUS solutions, I will compute BOTH methods for you (an LDP and a suggested modified SPC) and post (or privately Email) the results for your review.
My question about TOTAL relief is VERY important, and without that information, it's impossible to analyze your specific situation.