I have picked on Morningstar here and on RealMoney.com more times than I can remember. Greg's take seems to be similar to mine -- Morningstar's system is very close to worthless.
The big issue, as I see it (and Greg touches on this) is that the ratings are based on past performance. Morningstar has a lot of CFAs in its stable. Weaving together a reasonable forward looking analysis on a country, cap size or style is not that difficult to do. Clearly any effort to do this across the board would yield some conclusions that are correct and some that are wrong. If they are right more often than they are wrong they would be adding value.
Case in point: Awhile back I wrote a piece for RealMoney about Sweden possibly getting ready to have a run of outperformance. The forward looking analysis applied was quite simple. Higher GDP growth is expected (which is good for the economy) and the Riksbank (the Swedish central bank) was just starting a tightening cycle (good for the currency).
So far this turned out to be correct. iShares MSCI Sweden Index (ETF: EWD) has outperformed SPY by about ten percentage points and the dollar is down a couple of krona cents vs. the Swedish currency. I pay attention to Sweden on a regular basis, which is how I was in touch with the idea. I can guarantee that at some point the same circumstances will present themselves for some other country, and that trade will work too.
The point here is that forward looking analysis can does not have to be difficult. A firm like Morningstar could easily do some simple things like this and actually provide some useful insight.
To be clear about one thing, the circumstances that lead me to the Sweden idea simply put the odds in favor of the trade working. When I see something like in the future, I will probably write about it -- but the trade may not work. If you can put odds in your favor more often than not, you will probably do well but results are never guaranteed.