A good metric of battery life is the total energy it provides during its lifetime, which is
number of charge cycles*
capacity at full charge. If your phone uses X Joules a year and the battery is able to provide Y Joules over it's lifetime, it will last
Taking the data from Battery University, I came up with a following table showing the effect of the charge voltage, V on the number of change cycles, N, energy per cycle E (in %) and the total energy N*E:
Here, E*N represents the number of equivalent charge cycles of nominal (100%) capacity. As the charge level rises, there's a small increase in capacity, but it's dwarfed by a huge drop in charge cycles. For 62.5% charge (3.9V), the battery can produce an energy equivalent of 2000 100%-discharge cycles, 5 times as much compared to full charge. Here's a plot of the table above, for illustration:
It should be noted that this is not an original research uncovering a conspiracy of phone manufacturers. Cellphones simply are not expected to last more than a few years, and 400 cycles is plenty for that purpose. Batteries which have to last longer (e.g. in space probes) are using cycles with 30-50% depth of discharge.
Whether charging your phone to 60% of its capacity makes sense is debatable. If your phone has a small battery and needs to be charged every day, you simply don't have some 40% of capacity to spare, and having to charge twice a day is not practical. On the other hand, I have one of those Energizer-branded phones with a beefy battery, and charging it twice a week instead of once a week is totally viable.
Of course, there are other causes of battery ageing and failure than charge/discharge cycles, so the net impact on battery life will probably be smaller than this data predicts. Plus, cellphone charge controllers are designed to work with 100% discharge capacity and may work incorrectly if they hardly ever see the battery charged to 100% or discharged to 0%.