> You still need some code that parses the non-RSS part anyway, which can validate it.

Provenly incorrect.

How many RSS validators validates the contained HTML (as HTML, not as per XML well-formedness)?