For instance, supposedly space will end scarcity… and yet, any habitat in space will naturally have only a single source of food, water, and, even more urgent, oxygen, creating (perhaps artificial) scarcity.
Huh? Sure, if we forget absolutely everything we ever knew about reliability engineering.
Take air, for instance. If you're considering a community on the scale of a town or city, expect that it will be naturally divided into smaller physical units, corresponding to smaller social units in the community. Rather than having one big air supply for the whole "town" — which can fail or be sabotaged, creating an existential risk for the whole community — it'd likely be much safer to have small air systems for each household, neighborhood, commune, or other unit. You probably have to have them anyway for emergencies.
Infrastructure for distributing the air once it gets to the settlement is one thing. At least for now, though, Earth is the only place to get oxygen in life-sustaining quantities, which is the single source they're talking about.
Maybe you can posit harvesting oxygen from mineral oxides, hydrolyzing water if you can find it, or capturing an ice asteroid. Whether you split every atom of oxygen you breathe out of rust or lift them out of earth's gravity, let alone doing both for redundancy, it's orders of magnitude more energy and complexity than growing potatoes in Antarctica.
At least for now, though, Earth is the only place to get oxygen in life-sustaining quantities, which is the single source they’re talking about.
If we're talking about space colonization then "at least for now" doesn't apply any more.
There are vast quantities of oxygen available everywhere in the solar system. Extracting it is really not hard. There's a technology demonstrator generating oxygen on Mars right now. If you're arguing against space colonization because you're assuming that every bit of resources the space colony uses will have to be sent there from Earth, you're completely missing the basic concept of space colonization.
I often find that technological pessimists are imagining some very specific flawed scenario and then arguing about how that scenario is terrible or impossible, rather than arguing about the technology in general. Often the best way to debate them is to start by getting them to clarify exactly what scenario they're thinking of, the limitations of their argument will usually become quite apparent just by doing that.