This article is conflating all kinds of things. It says traditional cooking "outperforms" modern but is using a metric that essentially nobody cares about (thermal efficiency). It's just motivated reasoning to argue against electricity.

> The main discrepancy with these figures is caused when one doesn't take into account that electricity first needs to be produced in power plants which sometimes convert less than a third of the primary energy into electricity

That is, induction stoves are about 90% efficient at converting electricity to useful heat, but your electricity generation may not be depending on where you live. It's the same argument people make against electric cars and heat pumps.

As the grid decarbonises, electric cooking will become far more favourable than combustion cooking in terms of CO2 until eventually, hopefully, it is entirely zero emission.

Indeed the article doesn't seem to focus on what could have been its strongest argument: wood/biomass already IS carbon neutral, provided it's sustainably farmed. That does, however, exclude the fossil fuels used in production and distribution which may be significant or zero depending on source, but are usually quite low, albeit human resource intensive. If you want to go one layer deep you should aggregate up the carbon output of all the labour going into the production system but I digress.

Thermal efficiency only matters if you consider the carbon intensity of the source of the energy.

Besides, the major problem with combustion cooking is the particulates anyway, which are immensely better on an electric stove. And the article seems to consider convenience as an irrelevance rather than the primary motivating factor of essentially all technological development.

Unless the electricity comes from renewables...