Kerosene (and liquid petroleum-derived fuel generally) works well as an aviation fuel because it is relatively stable and energy-dense in both volumetric and mass terms. That it often has a high EROI means it can be dollar-cheap as well, which doesn't hurt, but it's still used even when the EROI is terrible (eg. military aviation - delivering fuel to a war zone costs a lot of energy) because the other advantages are still overwhelming relative to the energy cost.
> Hydrogen and e-fuels involve substantial energy losses even in ideal circumstances
Absolutely, though this is true of every industrial activity.
> and require dedicated supply chains that would not otherwise exist, which adds additional energetic costs.
While this is certainly true for hydrogen, e-fuels (which I take to mean synthetic electrochemically-produced hydrocarbon fuels, though perhaps you mean something different) can be produced and distributed using the much of the same infrastructure that exists today - for example, if you're making Sabatier methane, there's no reason you need to build new pipelines and tanks for it, though I'll grant that you do need new production facilities.
Of course, the rub is that the energy to drive it all has to come from non-carbon-producing (in operation) sources - solar, wind, hydro, or nuclear - and there's no denying that building out that infrastructure is going to be a massive project.
> In a world where available energy is going to become scarcer, I don't think using these fuels is going to make sense.
Available fossil energy is indeed going to decrease, but what makes you think that's true for renewables too?
As I was writing this comment, I took a small detour to do a bit of rough figuring here:
- Assume $77/MWh (Q1 2021) for utility-scale solar+storage [1]
- Assume 75 cents/L (today) for wholesale gasoline [2]
- Assume 105 L of gasoline contains 1 MWh (roughly) [3]
Then: $77 of solar energy can, if 100% efficiency were possible, produce 105 L of gasoline having a wholesale value of $78.75 at today's prices.
Obviously, 100% energy efficiency (or anything close to it) is impossible. But if solar energy instead cost $30/MWh, it suggests that an industrial process with an overall efficiency of 38% in energy terms (i.e. 2.625 MWh is needed to make 1 MWh of fuel) would be able to produce gasoline that matches the current market wholesale price.
Unfortunately I wasn't able to find anything quickly that describes industrial process efficiency in quantitative terms, so I don't know if 38% overall efficiency is reasonable to expect; personally, I think that probably would be considered high efficiency, so solar would need to get even cheaper (which it will) and/or gasoline will need to get more expensive (which it will) to improve the economics.
[1] https://www.nrel.gov/docs/fy22osti/81325.pdf
[2] https://markets.businessinsider.com/commodities/rbob-gasolin...; price quoted in gallons and converted
[3] https://hextobinary.com/unit/energy/from/gasoline/to/megawat...
EDITED because I can never remember to double-space when I want newlines