[css-ui-3] reorder sentences, reduce duplication in transition request

Author: frivoal

css-ui-3/transition-pr-2017-11.md

@@ -29,20 +29,18 @@ First Tuesday or Thursday after a sucessful transition decision.
 [Changes since the 2017-10-25 Candidate Recommendation](https://drafts.csswg.org/css-ui-3/#changes)
 <!-- no substantive changes -->
 
-The following changes were judged by the CSS-WG to be within the bounds allowed for a CR->PR transition:
+There are 4 normative changes, all based on feedback from actual implementations on a few subtleties of how they had implemented the features.
 
 * Clarify (as a SHOULD) the implications of text-overflow on pointer events to capture implementor consensus (corresponding test).
 * Clarify that UAs may ignore the cursor property to reflect the UA’s UI state
 * Allow, but stop requiring support for SVG images without intrinsic sizes for cursors (corresponding test update).
 * Align the spec with implementations, and make cursor: auto look like text over selectable text, and over editable elements (corresponding tests).
 
-These are normative changes, but they are all based on feedback from actual implementations on a few subtleties of how they had implemented the features.
-
 The first 3 changes are are small tweaks and cannot make an implementation that conformed to the spec prior to the changes become non-conformant.
 
-The 4th change could theoretically impact conformance, but in practice does not. 
-Zero known implementations (Gecko/Edge/Webkit/Blink/Presto) conformed with the previous wording (the [issue 1598 was openly reviewed and discussed](https://github.com/w3c/csswg-drafts/issues/1598) for several months), 
-while multiple implementations pass tests and interoperate with the change, 
+The 4th change could theoretically impact conformance, but in practice does not.
+Zero known implementations (Gecko/Edge/Webkit/Blink/Presto) conformed with the previous wording (the [issue 1598 was openly reviewed and discussed](https://github.com/w3c/csswg-drafts/issues/1598) for several months),
+while multiple implementations pass tests and interoperate with the change,
 and all others for which conformance was not affected (did not pass before or after) agreed to update to conform to the new wording (over the old).
 
 Thus the editors and WG believe all four normative changes fall within the bounds allowed for a CR->PR transition.