Marshac - good hunting, thats a decent explanation of it and what you should expect.
By the way, as explained in "The four pillars of investing", the gordon equation is not only good for the US stock model, it works on just about every prior civilizations "equity market".
Bernstein uses an analogy in the 4 pillars that I like and use often. A man walks a dog from his apartment to the park. The dog runs back and forth during the walk. The mans path is the gordon equation. He's going to the park. He's gonna get there. You can bank on it. The dogs path is the day to day movements of the stock market. Its the speculative (and unpredictable) component of the market. Indexers say they buy the total market and go fishing, and let the gordon equation pay them off over the long haul without worrying about where the dog will go tomorrow.
Make sure you also read this one (which was pointed to from the first article):
http://news.morningstar.com/doc/article/0,1,114455,00.html
Note that they expected return range for the next decade or two (2.1-4.2%) is before inflation, and almost completely consumed by it.
This is what you're going to get, in aggregate, from the broader markets over the next 10-20 years unless PE's go higher (and they're already more than 50% higher than average), or earnings rates accellerate to a rate higher than any time in history.
Now someone needs to jump in and explain how we'll get historic return rates 'because thats the way its always been and the future will be the same', when all the other parameters are banging their heads on levels that have no historic analog.
You're stuck with the paradox of "its not going to be 'different this time' because its going to be different this time."
If you buy into the historical mathematical model, you have to buy into what the same models tell you is going to happen going forward.
This is the thing thats bugged me about index investing from the get-go. If you buy into indexes at the point where they're at or below the "gordon line", long term over 20+ years you're going to make a very predictable amount of money. If you buy in above that line, especially if you're well above it, you can count on making nothing or losing money over 20 years.
I think it DOES matter when you get in. Or get out.
But you can wait a long time before seeing results. Clearly the market looked overvalued against the gordon equation in the mid 90's. But it kept going higher. You can lose a lot of money waiting for "the right time".