is my thinking wrongheaded do you think ?

"Wrongheaded" is pretty harsh : )

I just find that the best portfolios, even those needed for current income, are usually constructed with security selection done ignoring dividend yield as a criterion. Even after tax considerations, usually. My advice: Build the best portfolio you can, have a look at what yield it's getting, and if necessary sell a bit more to top it up to the income you want/need. In any case, it's unlikely in the extreme that any dividend slate will have the precise amount of income you want/need anyway, so you still have to fiddle a bit.

Would you rather have something with an earnings yield of 4% and a dividend yield of 3%, or something with an earnings yield of 8% and a dividend yield of zero? The answer is obvious to me (the latter), but some people will never see it that way. They think of a dividend as a costless something "extra" on top of the stock price return, so they unwisely want as much as they can get.

A portfolio of all stocks in the Value Line set that pay over 6% dividend yield underperformed the S&P 500 in the last 20 years. It seems they are bid up to overvalued levels by those that unwisely seek current yield. On average, this effect hits all stocks with meaningful yields to some extent. A lot of managers opt for a juicy dividend when they have no other good way to attract investors.

Here is a very different suggestion: a broad slate of large cash-rich dividend payers with high ROE. (version 2 in the post).
http://www.datahelper.com/mi/search.phtml?nofool=y...
That slate, 40 stocks held two months equally weighted and repeat, has beat the S&P by just over 6% after trading costs in the 40 months after that post.
And every stock has a dividend. The aggregate yield probably isn't too bad (it was 2.2% at the time of the post), and the risk has to be pretty constrained. Much lower concentration risk than the S&P.

Jim