Pitching Problems

I have recently thought how far are we in the field of baseball research or saber metrics?  Hitting I would say has progressed a lot, and so has fielding in the last decade or so but what about pitching?  The recent stats for pitching have been half fielding, regarding to DIPS and FIP.  Both of these attempt to take out the element of defense.  But what about the rest of the pitching field?

If you just answered W-L, then you are not reading you're type of blog.  The W-L is a terrible stat for the following reasons which I will repeat.

 1.  The pitcher must pitch well, allowing few runs.
2. The offense must score enough runs.
3. The defense must field all batted balls hit to them.
4. The bullpen must hold the lead.
5. The manager must leave the pitcher in five innings, and not remove him if the team is still behind.

Only one of these is in control of the pitcher, meaning simply W-L is just not a good stat at all. 

Now, after throwing out W-L, what would be you're next answer?

ERA?  ERA is a decent stat, but as we should know it does have it's flaws.  1) It shows the performance of the pitcher and his defense.  2) ERA can look shiny, but only with luck, defense, large strikeout numbers, and few walks.  ERA should not be a stat that you can rely on to tell you how good a pitcher really is. 

Pitcher's WAR?  Possibly, but as I believe I have said before why replacement level?  Why not just W, that is not compared to replacement or average or whatever else you could look at it. 

Those are the three I will talk about, I know there are many more but I think I've got my point across, when will somebody try to make a great pitching stat?  I want to, but how?