Seems like an extreme position to me. I propose that game should be well enough designed that at least a good portion of the people are motivated to play better and learn to master the game by the promise of getting whatever content it is that is "locked" behind difficult segments. What's wrong with that?
And remember, there's always Let's Plays for the casuals who don't want to get good at the game but still want to see it the final content.
Most people only play a game once, heck, most people don't even beat the entire game in the first place.
I don't have a problem with the argumentation that you'd want a game to motivate the players to better themselves. But what you're asking here is completely different, you're asking a player to essentially replay the entire experience (my in game clock in Isle of Rebirth is sitting at 25 hours, for reference, and I'm not at the end yet), just to experience that last bit of content. To be honest, I think most people will feel a bit disappointed at that and have no interested in replaying the entire game just to see the extras.
I don't think Shanes statement is extreme at all, if content is cut from the easier difficulties than 'is this game worth playing on anything below normal?' is a very real question that players are gonna ask themselves. And the end result will probably be that many people will play through a quest on a difficulty level that they are probably not comfortable with. I mean, heck, I have done this myself what with Isle of Rebirth. There is extra content in that which you can't access if you go down to easy, so I have not really had any interest in dropping down, even though I find the difficulty on the quest to be fairly absurd.
I'm going to get on my soap box for a minute here. Undertale did one thing very, very right - it made the world react to the player's choices in a logical way, questioning the stale mechanics of the RPG genre. However, the changes in the world are, for the most part, fairly superficial. Yes, you fight different enemies, and yes, they say different things, but the game itself doesn't fundamentally change. You don't have to find different solutions to problems and you don't get faced with new challenges based on how you've decided to play. It all breaks down to the (really awesomely fun, don't get me wrong) RPG battle or conversation mechanics. i don't blame Toby at all for not getting more wild with this; he said what he had to say, did something revolutionary and did it well, and then stopped within his limits.
This statement is fundamentally untrue though. Undertale is somewhat of an interesting game because it does so many cool things, but at the core it mainly have three different play modes that are based on player actions. And there is very much so content that is locked to specific modes. One mode features two completely different and new boss fights, while the other two has the same bosses, how you beat them are very very different between the two. One of the modes as a completely unique area tied to it.
I also don't think you should understate the plot, the changes in the narrative between the modes are fundamentally different, and it certainly will have a large impact on players who play for the plot.
If anything Undertale would support the argument that locking content behind different things works, but then, the different modes/paths in undertale are not really that different in regards to difficulty, and the game is actually cleverly designed so that the player won't have to face the harder stuff unless they have already done the easier.
But what if, instead of a game that tweaks the boss battles depending on your choices, -- or just gave bosses more less HP depending on your menu selection of "easy" or "hard" -- your choices actually make it into a different game? Let's say your goal is to rescue the princess. Perhaps you go in, sword swinging, and slash your way through the guards in the usual manner. This leads you into further gameplay of that kind, probably considered the easy-mode. The game notices you're a fighter and so the in-game scenarios and characters adapt to that. Or, perhaps you find a way to sneak into the castle and dodge the enemies instead of stabbing them, leading to a much harder, stealth and puzzle based challenge. Or, maybe you decide to murder not only the obvious enemies, but also all the innocent townsfolk and bystanders, leading to a game where EVERYONE wants to stop you and you don't get help from shopkeepers or anything. Maybe this is harder? Maybe it's easier?
Different choices lead to different difficulty levels which lead to different games. That might be a way to suit everyone's tastes. Of course, if you were going to do this, you'd need to have some kind of game structure that limited the amount of branching possible to keep the project manageable. Hmmm...
While interesting, and I do think it's a worthwhile concept that most certainly works, it's also not really very relevant to this quest. Such a large design choice would have had to be made at the start or very early in production, and it makes for a fundamentally different type of game.