Given that bold part, it all seems a bit moot to me.
I'd say there are at least some things we ought learn from solo instances, but, honestly... they're usually pretty distinct from anything you'd take a group to. The "challenges" they present are pretty well absent from group content, except perhaps stack markers still meaning "stack here" as much as they do anywhere else, and even at normal difficulty the challenges we face in group content are largely absent from solo instances if only because the latter is so leniently tuned. By the time it could actually teach anything of value, without being specifically reshaped to do so (which I'd be fine with, but is a fanciful notion at best for now), it'd feel out of place yet again -- just, in the other direction.
I guess all I can hope for is to see some revised guildhests that are worth doing despite continuing a difficulty curve upward, that leveling dungeons can be retuned or revised slightly to have at least some bite -- and increasingly so over time -- and... nope, that's pipedream material again already.
I'll just say this, then: there are opportunity costs on both sides. What people do in their solo time can matter to you when you later group with them, in the same sense that whether your classmates in a discussion-based class have done their homework matters to your learning; you can't force anything, and it's probably not worth doing anything about it specifically, but it is a factor when you'd like to see [A] and are therefore willing to do [B] while others are unwilling to do [B] and therefore will keep you from seeing [A]. And of course the reverse is true as well. If most people want more content that doesn't feel like a joke, what they want will slate you for having to improve a bit more. Those interests may conflict. The trick to design is then making them conflict as little as possible, not by splitting the two groups away from each other and giving them each their half, but reducing the irritations of reaching and playing among that average level while increasing the enjoyment that can be gleaned even when playing content made for that average and expanding the number of playstyles and difficulties the content can accommodate. Some things will still be split, of course, but that shouldn't be the first tool design reaches for.