Decoding is not a science; (what did the guy with a hammer and stamp do, when, where, and what stage).
I had a Snyder-Rollfast bike, with both a "35EH" stamped bottom (i.e., typical 1953), and a K-serial i.d.; so my theory would have required a skipping of both I&J; [but that would contradict a J-serial i.d. Snyder-Hawthorne]. Also had a post-war 1949-E that seemed to fit in line.
So, as with other manufactures, perhaps bottom brackets were stamped as separate parts, thrown in a bin, and later used with a frame, in an assembly step. Thus there may be overlaps of various codes, dates, and stamps. What I try to do, is find a date that I feel comfortable with, based on limited information, and if upon receiving better info, revise my previous assumptions.
For example, I was given a road bike with a documented December 1985 frame serial number (5L), with documented features of the 1986 catalog (Imperial Red paint). So, I call it an 85, 86 or 85/86 - close enough.
And then there are sometimes those odd, unique, or not understood markings, (such as your "473C").