Dividend omission is the biggest enemy of income investor. There are two types of dividend omissions (and cuts) - predictable (for companies with pure financial situation , e.g. after few years of negative earnings) and non-predictable (e.g., for companies in good shape) that can be viewed as Black Swan events.
Many dividend growth investors (DGi) have quite concentrated portfolio from 5-12 stocks. Probably this is OK if you are perfect in stock selection and can closely follow market, so you can react within few minutes (in our days of high-frequency trading probably this period squeezed to few milliseconds) on company announcement about dividend omission.
Based on normal distribution of stock prices (this is not so good approximation of reality) economist usually suggest diversification with 25-35 stocks, i.e. invest in each company 3%-4% of money. While average DGR ~ 6.5% and average inflation ~ 3.5%, average real income growth rate ~3%, so dividend omission in 1 of 33 equally weighted companies in a DGi portfolio is quite significant event. Unfortunately average dividend cut probability (~4%) that can be a Black Swan event is higher than estimated average real income growth rate (~ 3%).
Because of market uncertainty a DGi can be sure only in one number - his (her) yield on cost (YOC). Dividend history, DGR, number of cuts, etc. are all from history. If DG investor plan to sell some stocks (BTW, /s/he must do it in regular IRA at certain age) - market neutral portfolio (stocks with anti-correlated prices) is a good approach but again all correlations (or beta) are from history. With historical average probability of dividend omission ~ 3% (assuming that stock price also drops to zero) it is better to keep each position below this 3% and have yield above this 3%+inflation rate. So you face 1 omission if you hold 34 average dividend stocks per year. If you hold 11 stocks you face 1 dividend omission in 3 years.
Because dividends cuts are strongly correlated within a single country (at least USA, UK, Japan) and cuts are often 50%-90% DGi should held about 2X-3X more stocks than "regular" investor. On another hand, because standard deviations of DGR for stocks is significantly smaller than stock prices standard deviations. Textbooks recommend that good equity portfolio should consists of 20-50 stocks (20 was recommended while ago when price correlations of stocks were relatively small, 50 is common current recommendation). Textbooks recommend that good bond portfolio should consists few bond funds because authors of textbooks consider that small investor cannot assemble good bond portfolio him/her-self. One quite good book I read (sorry do not recall the title) recommends that good bond portfolio should consists about 500 different bonds. Because DG stocks have non-zero DGR in contrast to bonds with DGR=0, I think that DGi should have smaller portfolio that a bond investor.
Also from history of US stocks probability of dividend omissions and cuts for DG stocks is lower than for other dividend stocks, so numbers are slightly better. Nevertheless, I think DGi should have wide equally weighted portfolio from 100+ stocks.