You can read that "downside" more generally as DAC (and carbon sequestration in general) is expensive, and no one really knows how efficiently we can do it, especially at the scales we'd probably want.
It's also an expression that current low/no carbon energy sources are globally sufficiently constrained that we're in no real position for large scale (aka meaningful) DAC (and many other forms of carbon sequestration - perhaps not all) deployment.
Climeworks is clearly out to iterate and gain experience to be well positioned for a presumed DAC (and sequestration) market that may form as all of the easy to decarbonize sources and processes are converted. We're... a ways off here.
This is addressed in the article: "Or would the geothermally-generated electricity go to better use powering electric cars?"
There's no such thing as waste heat really. In the short term you can argue that limited units of energy are better used towards carbon neutrality, things like powering electric cars. Ultimately, we need removal tech, and we need to make it energy efficient, fast. Putting money and energy into carbon removal systems needs to be looked at as an investment for it to make sense. Powering an electric car gets you to go a mile, powering a DAC system lets you remove a little carbon and is an investment in developing more efficient systems too.
If you have anyone in mind to reach out to, climate scientists I know love to answer curious questions like this. AirMiners has IPCC co-author Bill Collins speaking a few weeks, you could see about posing the question to him: https://lu.ma/airminers
What are the actual downsides of energy usage here given that the enterprise is strongly carbon negative? Given that it is consuming geothermal energy in Iceland, 1) wouldn't the energy become waste heat in the environment regardless, 2) Is there a consumer of the energy that would be "better"?