They are semantically simple string operations whose computational complexity scales with string length.
Near unbound string lengths would exhaust the time (somewhat aproximated by instruction) budget of eBPF applications doing even a single one.
I think the point was not to show "there are a lot of helper functions" but rather to point out "the (undocumented) helper functions that exist are problematic"
