To comment on my own comment, what we are really arguing about is the form of the frequency distribution of suicide ages T_ suicide conditional on cheap, readily available anti-ageing tech.

The OP is basically saying that that distribution will be close to frequency distribution of age at death for ageing-related deaths that we see today (see page 9 of that document)

I am saying it will look nothing like that; the psychological factors determining suicide decisions will just produce loads of variance. It'll look more like a lognormal, where the log10 of T_ suicide is normally distributed with mean 3 and variance 2. Actually even that is probably wrong, but it's a pretty good starting point.