A 3-D characteristic of retirement income portfolio withdrawals means that there is more than ONE point in such a 3-D matrix.
What does that one point suggest? Just ONE time period (typically 30 years*) AT just ONE allocation (typically 60/40*).
*Typically meaning the common research and discussed variables for retirement income withdrawals.
But what about the other allocations? Especially those other allocation choices as a retiree AGES into those ever shorter time periods? In other words, both allocation and time periods change when viewed through a 3-D matrix developed through running multiple comparisons between both changing allocations AND changing remaining expected longevity (i.e., an 80 year old does not have the 30 year expectation a 65 year old may have).
Modeling aging needs to consider longevity effects on time period expectations combined with allocation adjustments as a result of that aging.
Most research today takes an ageist view on this question, meaning advisors and researchers are biased by their "young" not-yet-retired age that influences allocation and time periods they research. People don't stay young in retirement - they continue to age.
How do longevity statistics change as one ages? Graph illustrates the aging effects on time periods.
Here is research unincumbered by such views.