Consider for your acceptance criteria:
"All linear indications over 1/16" are to be reported. Cracks, laps, and seams are rejectable." This gives you a documented way of accepting the inclusion stringers that will meet your specification. If inclusion stringers are not acceptable for your product, you will have to use a higher quality steel (I would doubt that inclusion stringers are injurious for your product).
Are you dealing with new, freshly machined shafts, or inspection of used shafts? Inclusion stringers, which can produce an observable indicaion on a freshly machined surface, are not as detectable on "aged" surfaces. A lot depends on surface finish.
One of the problems is that the inspector is not going to be able to determine what causes the indication, only its location and orientation. I've seen PhD Metallurgists, after examination with optical and electron microscopes, argue whether or not a specific imperfection was a lap, seam, quench crack, or SC crack, so it is not realistic for a Level II inspector (with what, a HS diploma and a combined total of 48 hours worth of training?) to be able to accurately identify the source of any indication with just MT. A Level III should be able to opine on whether or not a specific MT procedure is capable of detecting these imperfections, but he won't have any better idea than the Level II (or the PhD's for that matter).
However, not many inspectors will argue whether a "crack, lap, or seam" is actually a crack, a lap, or a seam. You may get some that would argue that they can't tell a stringer from a seam, but I'd say those are likely ones that have not seen indications from stringers and/or seams.