DCA is not an answer to what happens to the lump sum that is already invested, and mainly only applies for people who have a very long time horizon for which they will keep adding to their accounts - i.e. the future cashflows to be invested are >>>> than current account balances
Re dividends, if you are long stocks now in the above mentioned accounts, seems to me the < 2% dividend yield of the market overall doesn't even keep up with inflation
I believe you are looking at it strangely. I know of almost no one who's put all their money for retirement into the market at one single point. Virtually everyone saves for several years (or decades). This is by definition DCA. Ergo, you cannot take a market peak and look 20 years down the road and say "well, their money didn't increase in value". It did increase, on average, compared to the amount they put in.
Also, the chart you posted (which was very slanted) highlighted three periods, each >20 years, where the SP did not make any gains
without dividends invested. 20 years to me is a long time frame. In truth, no 20 year period in the last 130 years has seen a negative real market return for the SP500 with dividends reinvested. Ever. Could it happen? Sure. But it hasn't yet - that's my basic point regarding that graph.
Also, if you're looking at dividends as being unimportant because "overall [they don't] keep up with inflation" i believe you misunderstand the power that they can have. The power of dividends isn't that they are on average just above inflation, but that they are gains on top of increases in share price. Better yet, they are their own kind of DCA because they are issued every quarter, whether the market is up or down.
here's a simple example. Over the past 40 years, the SP has returned on average 3.38% without dividends, but 6.48% with. An investment would increase 3.7x over that time period with dividends, but a whopping 12.4x with dividends. In other words,
dividends accounted for 70% of all returns.