http://finance.yahoo.com/q/bc?s=MSFT&t=5y&l=on&z=l&q=l&c=%5E...
If you were to solely depend on S&P500 as the index for performance, I guess investments in Lockheed Martin would be a waste.[1]
[1]http://finance.yahoo.com/q/bc?s=LMT&t=5y&l=on&z=l&q=l&c=%5EG...
There's an argument that dividend payers outperform the index over the long haul, but that's something different.
Person A has 100k to invest, and puts it all into Microsoft. He's betting MSFT will outperform the rest of the industry, and likely all industries. Presumably he spent some time and effort to research this decision, seeing as investing in a single stock is risky.
Person B wastes no time, makes no predictions, and just puts 100k in the general market. He gets the average appreciation and average dividend returns along the way.
Person B ends up ahead: less risk (one vs. many), better returns, and zero time in research.
You made some flawed assumptions:
[1] All companies in the index pay dividends.
[2] MSFT is not one of the companies in the index.
[3] There is only one industry in the index or all industries in the index are somehow complementary (ie. a loss/gain in one will see a similar loss/gain in another)
[4] Most industry are related to Technology (Only 62 out of the 500 companies are in the Technology sector)
Using your example, if the Technology industry outperforms all other industry within a given period, Person A would enjoy a profit whereas Person B might walk away with a loss.
Market indices shouldn't be used as the benchmark for the true reflection of a company's performance/worth. Lockheed Martin is the key example, its one of the most desired stocks that has been growing steadily and pays one of the highest dividends, yet it underperforms the market.
Taking a job to work at McDonalds is a bad decision when the alternative (picking a random job available to you, essentially) pays more. Even if you "get paid" while working at McDonalds, it wasn't a worthwhile decision.
(By the way, stocks are absolutely compared against the average market to see if they are a worthwhile decision. Why do you want to invest in an underperformer when you can invest randomly and do better? It's like testing a drug against a placebo.)
If the broader market returns 10%, you can't say, "But I got a 3% dividend, that's apples to oranges." In the end it's all part of the total return.
I'm not arguing that dividends don't matter - on the contrary they matter a lot. But a stock with low total returns over time isn't saved just because the dividend portion of that is high.
2) You can make a judgment that in the future this will change. What's the basis of this, when most stock pickers are wrong as often as they are right?