fix: normalize KD-CAL and Kind-CAL teaching layer for G as scope object

[iqdoctor]

Background

This PR addresses a residual post-#29 inconsistency, not a reopening of #29. It is limited to the KD-CAL / Kind-CAL teaching layer.

Problem addressed

Closes #35.

C.2 / C.3 still taught G as if it were a point-characteristic or coordinate-like member of the same genus as F and R, even though post-#29 FPF normalizes G as a USM scope object.

Minimal repair

• replace three point-characteristics / point-values framing
• redefine G in C.2:1 as applicability over U.ContextSlice
• rename the C.2:4.1 subsection so it no longer teaches coordinate-style grouping
• replace characteristic 1 (USM, G) with scope object 1 (USM, G) in C.3

Out of scope

• no changes to A.2.6
• no changes to B.3
• no unrelated cleanup outside C.2 / C.3