You raise valid points, but I think both are genuinely answerable.
Yes, they are answerable. But until they
demonstrate that they've answered it, there's no reason for confidence.
1)...
The first positive sign was with DNDBeyond, when the abandoned working with the previous (incompetent) company trying this and move to the Beyond people, who, despite some concerns about their pricing model, have actually put out a very good product with only a couple of serious issues.
If we had evidence that choice came from
understanding the domain, instead of just luck, that would be a good sign.
And since then we've seen a lot of successes - BG3 is looking good (the script has Larian issues, but that's par for the course), Dark Alliance is looking surprisingly non-awful (and is developed by a WotC-owned studio), and even if it's a bit mediocre, it'll be a stepping stone.
None of these are actual successes yet. They are projects that haven't yet failed. There's a difference. A success is good code that ships, not things that are not-awful in previews.
2) The product DNDBeyond is putting out is actually surprisingly close to a lot of stuff I work with, in that it's essentially a big-ass database which has to be very user-friendly and accessible, is commercial in nature, and has users viewing stuff as well as inputting different stuff, and so on.
Fundamentally nothing that they're doing there is hard or complex to do, nor is it expensive to run something like that.
So, you're not the only one who works in software around here, so I'm going to push back a bit on that - it is complicated and complex. Yes, many companies do similar work, but also many
fail to do it. Moderate data complexity plus moderate customization per user, millions of users, three-nines uptime and a reasonable UX design is not a cakewalk of a project.
The number one way to make sure a project blunders is to be overconfident about it.
On top of that, they have to go with WotC's release dates. You seem to be thinking of this like the WotC D&D devs will just go on their merry way, like the digital offering doesn't exist, that they won't give long-range heads-ups on systems, or even change systems, to make them work in digital.
As soon as you have two systems in active development become dependent on each other, things start to suck, fast. "You can't release that physical book that you've already sent off to the printer because we've got persistent issues implementing a class on the software side," is
NOT a conversation we want them to have to have.
I think I would prefer to not see the software side have a lick of input into the RPG game development. I am still more than willing to use paper and pencil, so I want game developers to go with whatever works for the game, without concern of how hard it may be on the software side. As soon as they are linked, you're apt to see game design hobbled to the digital implementation.