> The two big non-electrical energy demands are transport ..
A large fraction of transport is not amenable to electrification in this manner; however transport is the low hanging fruit and I support the rapid electrification of transport where possible. Unlike batteries for cars, generation of biofuels/hydrogen for airplanes/ships/heavy trucks will not be significantly more efficient than fossil fuels - it likely will consume more energy, not less. The fossil fuel technologies are already very efficient (50%+) and renewable alternatives are very inefficient. It is possible to electrify/solarise these processes in the long term, but also complicated and capital intensive. I have worked on technoeconomic simulations of such processes, where I learned first-hand from expert researchers in this field (though I am not a chemical engineer).
> .. and heating. For heat it just comes down to heat pumps
This is not true at all, and you've likely misunderstood what heating means in energy breakdowns.
Heat pumps are most suitable for low temperature heat - municipal heating, (industrial) cooking, etc. Things which are already largely electrified in developed countries. But low temperature heat is widely available as a downstream by-product of higher temperature processes (including power generation as in CHP), and it is there a low priority in the scheme of things.
Heat pumps are not feasible (nor are they even theoretically very efficient) for high temperature industrial processes, of which there are a great many (concrete, bricks, metal processing, plastics and other chemical processing, etc). Many of these processes are already practically 100% efficient, so electrification will use at least the same amount of energy. A factory may use for example a steam turbine with a mere 5% electrical efficiency - the high temp steam is used to heat a chemical reactor, and the small electrical output is used for pumps, etc.
Finally, the direct combustion of fuels, (often bundled under 'heating' in stats), also includes the use of fossil fuels as a chemical reagent, primarily carbothermal reduction of metals (plus many petrochemical processing reactions). This usage is highly efficient, and cannot be replaced with electricity directly. Alternative processes will likely consume more energy not less - there will be additional intermediate processes, separations, and so on, likely requiring melting/dissolving/reacting the materials at high temperatures.
