Refactored GlassBrowserFrameView and BrowserDesktopTreeHostWin.

chrome/browser/ui/views/frame/opaque_browser_frame_view.h

Index: chrome/browser/ui/views/frame/opaque_browser_frame_view.h
diff --git a/chrome/browser/ui/views/frame/opaque_browser_frame_view.h b/chrome/browser/ui/views/frame/opaque_browser_frame_view.h
index e70e8b4ede79d2c9f6a9f2158a3ca52241c2f317..49d20af02c54d32a03a1ca176806c8e67c06a201 100644
--- a/chrome/browser/ui/views/frame/opaque_browser_frame_view.h
+++ b/chrome/browser/ui/views/frame/opaque_browser_frame_view.h
@@ -126,8 +126,9 @@ class OpaqueBrowserFrameView : public BrowserNonClientFrameView,
   // Returns true if the specified point is within the avatar menu buttons.
   bool IsWithinAvatarMenuButtons(const gfx::Point& point) const;
 
-  // Returns the thickness of the entire nonclient left, right, and bottom
-  // borders, including both the window frame and any client edge.
+  // Returns the thickness of the entire nonclient left, right, top, and bottom
+  // borders, including both the window frame and any client edge, but excluding
+  // the titlebar/tabstrip area.

[Peter Kasting, 2016/04/15 00:20:52]

Now I'm really confused -- I don't know what "the entire nonclient top border,
but excluding the titlebar/tabstrip area" actually includes.

And I don't think this is ever used for a top border height, nor would it be
accurate for that (since, even excluding the titlebar/tabstrip, there's no
client edge on the top border).

So it seems like the old comment was more correct.

[Bret, 2016/04/15 23:07:08]

Hmmm okay you're right, this was the same mistake as the corner width comment
edit I made in the other file, I mixed up how the parameters of
GetHTComponentForFrame are used. Anyway, reverted.

   int NonClientBorderThickness() const;
 
   // Returns the bounds of the titlebar icon (or where the icon would be if