As many people have pointed out, an equation editor without *Tex is nearly useless except for trivial use. On the other hand, apparently this feature was removed in order to make life easier for beginners. I understand this motivation, as most users are not sufficiently knowledgeable to take advantage of Tex, and the graphical editor requires almost no learning curve for new users. However, perhaps it is possible to get the best of both worlds without changing the easy look and feel of the new WYSIWYG editor.
An alternative that would make the new equation editor usable by serious users would be to make the clipboard operations convert to and from Tex. There would be no downside to this, since the current clipboard conversion to text is useless (only raw characters are copied). What I have in mind is that you would be able to copy and paste latex code to and from the Google equation editor even if the GUI only shows the current simple editor.
This clipboard feature would require absolutely no GUI changes, so it should be pretty easy to implement, and would not inhibit beginners.
It would be better, though, if the graphical editor accepted Tex compatible keystrokes (e.g. semantics for curly braces), in which case pasting text content into the equation editor would be the same as typing the same text.
Another idea: allow the user to select an encoding in options to specify the translation from plain text. For example, the Mathematica encoding would facilitate integration with Mathematica. Just copy a Mathematica expression to the clipboard, insert an equation in Google Docs (which would put the user in the new graphical equation editor), then paste, and voila, your formatted equation appears, ready for further editing.
The transition to the new editor is worrisome because it seems to indicate a change in Google philosophy to something a bit more Microsoft-ish. I hope this is not the case, but it seems as if instead of listening to the opinions of Google's fabulous base of technological know how, apparently a few management executives must be making final design decisions behind closed doors based on a superficial analysis of usability by the masses. I'd find it hard to believe that the brilliant engineering base at Google would have supported the removal of the Tex based equation editor without noisy objection. I don't think removal of key features in the name of simplicity is a very good idea most of the time.