thought it would have checked for encoded characters and skipped them?

That would be BAD! Decoding and encoding your post would change "such as £" to "such as £".

On a side note just noticed something annoying about HTML::Entities, if your input is already encoded

Duh?

The same thing happens if you try to encode characters twice.
The same thing happens if you try to encode URL characters twice.
The same thing happens if you try to zip a string twice.
etc