The so called codes you refer to are way behind the learning curve on the state of the science and art of authentic codes.
They aren’t remotely long enough to be valid—i.e. short enough such could be found by chance in a sufficiently lengthy text.
They aren’t centered in a complex cluster of codes including at least a very long one==one or more of which cross significant keywords in the surface text.
They don’t have obvious and clear references to the surface text like the authentic codes do.
etc.
etc.
etc.