If we're talking about cycle life, calendar life and temperature, for lithium titanate batteries, then it's worse than that. You only get to pick one
It's something like:
If you keep them at 25 °C and don't cycle them (or cycle them infrequently between 20% and 80% SoC) then you get 10 years.
If you keep them at 25 °C then you get 10,000 cycles, provided you do it in within about 2 years.
If you keep them at 50 °C then you'll only have about 1/6th of the cycle life or calendar life you would have got at 25 °C.
[Edit: I remind you that this is for LTOs from reputable manufacturers willing to put their name on their devices, and willing to admit they are LTOs.]
The reduction of lifetime with temperature is based on Arrhenius' law
which applies equally to capacitors and batteries. It says that life approximately halves for every increase of 10 °C.
If instead we compare the properties of supercaps and batteries (of various chemistries), something like your diagram applies to the properties: long-lived, energy dense, power dense. Cost is remarkably weakly correlated with any of those properties, given enough time and mass production. In the limit, the cost tends to depend on the abundance of the raw materials used
, and the energy required to extract them and combine them.
Lithium ion batteries have high energy density, moderate life and moderate power density. Traditional (EDLC) supercaps have high power density and long cycle life, but low energy density.
There's a fundamental physical reason, in fact a simple geometrical reason, why supercaps are unlikely to ever have energy density (energy per volume) approaching that of batteries. That's because capacitors can only use the surface
of the material, where batteries can use the entire volume
Using the common analogy where water represents electrical charge, consider a stack of sheets of paper versus an equal-volume stack of sheets of plastic. The plastic sheets are of a kind whose surface is wettable (capacitor analog), but the paper allows the water to penetrate its entire volume (battery analog). The stacked plastic sheets, like the stacked layers of graphite, are useless as they stand. You need to separate the layers in order for them to hold any water at all. But it's worse than that. You need to prevent them from re-stacking when the water is drained. So you need to randomly crumple every plastic sheet to prevent them from re-stacking. As you can imagine, the stack of crumpled plastic sheets (graphene analog) takes up far
more volume than the stack of flat sheets of paper, and still doesn't hold as much water.
We can trade off power density for energy density by making a hybrid between a capacitor and a battery, by mixing their cathode materials, but the battery-like cathode material still has much the same lifetime it had when it was in a pure battery.
The thing is, lithium ion batteries already have more than enough power density and lifetime for electric vehicles, and even more so for home energy storage systems. We don't need the power density or cycle life of supercapacitors for these applications. We just need lithium ion batteries to get cheaper, as they are doing. And we need to research batteries that use only the most abundant elements, as we are doing.
I note that "end of life" is defined as when the energy storage capacity falls to 80% of the rated capacity, or the internal resistance doubles. In future, a battery that has had a full life in an electric vehicle may have a second life in a household energy system.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).