That's exactly why the inverted advice needs to be heeded, so that that sucky experience, which does still occur daily with a lot of sites and apps, goes away by default.
Note, I'm absolutely not saying that a big screen desktop experience isn't inherently better for all but the simplest apps, there's very high probability it is because it's more ergonomic hardware with far more screen real estate.
However since most people want to, and do use mobile in preference to desktop for a huge proportion of tasks now, the design philosophy in most cases needs to flip from 'full functionality for desktop, scale down gracefully for mobile' to 'full functionality for mobile, scale up for desktop, taking advantage of the extra UX potential where possible'.
It's actually a far more optimistic, and creative, approach. Make it, then enhance it for the less popular use case rather than make it, then degrade it for the less popular use case.
I strongly disagree. There's tons of people that talk about how everything should be mobile first, but that's just not feasible for a lot of web applications. Stop trying to shove mobile into everything.
If I'm making a Web IDE or a Web Photoshop, it's very unlikely I'll be able to fit all of the functionality that's needed into a tiny mobile screen, and it's also unlikely I'll be able to get it to perform well. And you know what? That's totally fine, because if my demographic is gonna be people with 1920x1200 monitors on a powerful desktop machines, it'll work great. I'll build an amazing experience for desktop, because that's my target demographic.
A lot of enterprise applications are impossible to scale down to mobile as well, due to the sheer amount of customizability and information they provide. I don't know of many enterprise applications that support both mobile and desktop. If you want to support mobile for an enterprise app, you're better off designing a separate mobile variant of your application. This assumes you have the resources to do so, and that there's sufficient interest from your customers such that the decision to have a mobile variant makes sense.
Here's the thing, building a sophisticated application that works well on a tiny phone and scales all the way up to a 30'' monitor is not feasible at all for a lot of teams. I'd challenge you to show me a good example of a sophisticated app (e.g. along the lines of a Web IDE or Web Photoshop) that will scale nicely from a tiny mobile screen all the way up to an awesome 30'' display.
... well, no. Technically you don't have to. But you almost certainly should.