Surely its taking income out of total return against dividends that is seen as being a different style/strategy. The HYP's look to achieve a dividend value that broadly rises over time. Total return can still be used as a measure, indeed its pretty much the only way to compare two dissimilar style portfolios to see how successful or not each style might have been compared to each other.
No matter what stocks you hold at some point likely they wont have done well over a 10 year/whatever period. If you buy stock and 10 years later with dividends reinvested they're down to 66% of the inflation adjusted start date value then spending dividends will have pulled that down further. 1965 - 1974, and 1999 to 2008 for instance. Equally other assets might see the same, or worse, gold 1983 - 1992, 1988 - 1997 for instance where gold ended at 45% of the inflation adjusted start date value.
If you started with 50/50 in stock/gold and just let that run, stocks accumulating dividends ...etc. then in each of the above cases the bad asset was more than offset by the other asset, that across each of those gained something like 2.7 times in inflation adjusted terms. So 50/50 ended up with stock light/gold heavy, or stock heavy/gold light, but where the combined value was of the order of the average of 0.66 and 2.7 = 1.7 times more in inflation adjusted terms (5.5% annualised real). Similarly however 50/50 stock/gold bought and held had its worst 10 year period, 1994 - 2003 where that ended with a few percent more than the inflation adjusted start date amount. These are all figures since the end of WW2 and where silver was assumed to be held up to 1971, gold thereafter (after the US ended the pegging of the US$ to gold).
Spending dividends, and in a bad case of where total returns with dividends reinvested over 10 years sees 65% of (inflation adjusted) portfolio value remaining is a considerable risk. Not considering total returns would hide that risk. In sight of the risk however you might deduce that it would be better to put perhaps 30% into cash/bonds for draw-down over 10 years, 3%/year and leave the stock 70% total return to hopefully grow enough to offset that and more. However in a bad 10 years with that 30% of cash all spent, and 70% of stocks having declined to 65% of the inflation adjusted start date value = 45% of the overall inflation adjusted start date value remaining after 10 years. Considerably less than the 72% that 50/50 stock/gold 70% with 30% cash had remaining in its worst case.
At what cost though? All stock in the average case near doubled after 10 years (1.97 times more in inflation adjusted terms total returns) whilst 50/50 stock/gold had 1.77 times more on average. Which on a annualised measure = 7% versus 6% type difference.
I very much suspect that such characteristics are evident in each of HYP, World tracker ...etc. choices and its important to understand those risks as they are not trivial risks.
If you start with 30% cash to cover 10 years x 3%/year of spending and draw that down to zero then it average 15% cash. Split the remainder 50/50 between stock and gold and compare (rounding) 16 cash along with 42% in each of stocks and gold ... compared to 16 cash and 84 stock and
reward wise they have been pretty similar since 1972 at least for US based data, but where the stock/gold choice did so with less volatility (which is commonly considered as being the better risk-adjusted reward).
Buffett advocates
90/10 stock/T-Bills, perhaps because he considers a 2%/year spend rate.