The description of this rule claims that
<em>have a semantic meaning whereas
<i>only convey styling information like CSS
This is factually incorrect since HTML5. Both
<i> do have semantic meaning and those semantics are different to the suggested alternatives of
<em>, respectively. There is no way of knowing whether the author is (incorrectly) using these elements for styling only, or whether they are correctly using them for their semantics.
By advising authors to switch from
<em> this rule is effectively advising to change the semantics of the document in a way which could be incorrect.
Moreover, by claiming in the description that the former elements have no semantic meaning and are for styling only, the rule is perpetuating a falsehood and miseducating readers.