Limited and constructed are different beasts when it comes to mana because the format is different not because mana works differently. A lot of limited drafting environments suck. And at least some of them suck specifically because fixing is bad. I don't see how this doesn't not only extend to cube but is magnified in cube because of how much higher power the format is.
What formats suck, specifically because of fixing issues? Is this a reflection on the importance of fixing specifically as something needed in abundance, or is it perhaps related more to poor limited design in general, wherein multicolor decks are made mandatory by poor set design, but unsupported by poor fixing options? At which point do we blame a limited environment being bad on exactly one aspect of it? In my experience, bad environments do not score 10s across the board, with one singular failing that brings the whole beast down. It's a constellation of bad, weird, or nonfunctional deployments in set design, extended to, not because of, the fixing base. There's a limit to how much we can learn from bad environments, and I don't think mana fixing is often their lesson, inasmuch as bad/boring/weak themes tend to be, at the very least.
Furthermore, I think it's useful to point out that there is only so much we can draw from Constructed and Limited as formats, because Cube is an entirely different beast. Most of us in no way curate a retail set limited environment, and that's a huge distinction that should be at the forefront of all of our minds. Retail set limited environments use a rarity hierarchy for the distribution of cards and have a great deal more room for weak draft chaff, while all of us typically strive for a tighter power band. So even if the claim that there have been limited drafting environments that sucked specifically because of bad fixing was true (which for the record I'm deeply skeptical of), it wouldn't necessarily have any bearing on cube, especially if said cube was built with those format constraints (less fixing) in mind.
The problem with running a thousand Mana Confluences is that not every deck can just take 1 every time they need a color. Evolving Wilds has a similar problem. Most decks can't afford to just have every fixing land ETB tapped. IMO, this is not a replacement for traditional dual color lands in a traditional cube list.
Very few among us run a "traditional cube list". Yes, Mana Confluence has downsides, as does Evolving Wilds. They introduce format constraints. I think that's kind of the whole point of their suggestion. After all, format constraints via manabase tinkering has proven to be a very cool space to explore (see: Grillo & Bouncelands, for one glowing example within our own community among many), so even if the idea does not slot cleanly into every list (see: Bouncelands), it at least provides an interesting new angle to take on cube development for those of us looking to shake things up.
For the record, I don't agree with the idea that adding in a bunch more narrow cards is necessarily the best way to capitalize on the space made by running less fixing, but I do think there's an interesting format just waiting to be cracked by pursuing this concept. I won't be cracking it anytime soon myself, but I do intend to heavily consider this as I continue to toy with my fixing base, though, and I think this posting was pretty timely, for me at least, as I was just thinking of how I wanted less fixing lands, and this has helped point out that there is an alternative space than the traditional double fetch, double shock, scryland model that's dominated Riptide for a while now.
edit: clarity