Remove special handling of <rp> in innerText's getter

This was a hack that is only implemented in one engine (Gecko).
For example, an <rp> in <select> is included, which is weird;
if an <rp> has several text nodes, only the first is included.

There is not enough implementor interest in making <ruby> actually
work well in innerText (e.g. in the absense of <rp>); see #1801.

source

@@ -11666,15 +11666,6 @@ interface <dfn>DOMStringMap</dfn> {
      'visible', then let the result of these substeps be <var>items</var> and abort these
      substeps.</p></li>
 
-     <li>
-      <p>If <var>node</var> is a <code>Text</code> node child of an <code>rp</code> element, then
-      let the result of these substeps be a single string item containing the <span
-      data-x="concept-cd-data">text</span> of <var>node</var> and abort these substeps.</p>
-
-      <p class="note">The <code>rp</code> element is normally 'display:none' but we want to include
-      its text.</p>
-     </li>
-
      <li>
       <p>If <var>node</var> has no associated CSS box, then let the result of these substeps be
       <var>items</var> and abort these substeps. For the purpose of this step, the following