That's a reasonable point of view.
My point of view is that we already call for an attack roll, and that can resolve whether or not the attack works.
If you limit potential actions by creating a list, players must consult the list before they attempt anything. In AD&D this was "I can't climb, I'm not a thief"; in 3E this was "Do you have the feat for that?"; and in 4E it was "Do you have a power for that?"
Not that there's anything wrong with that. But it means that the way players make choices is going to be different.
So we ask: what do players do in this game? Do they pick from a list of possible actions and execute the rules that relate to that action, or do they come up with an action on their own and apply the general action resolution rules to resolve it?
I prefer the latter because I think it draws on one of the strengths of RPGs - so-called "tactical infinity". (You don't really have infinite options because you're limited by what's plausible in the setting, but you get the point.)
I personally think tactical infinity is too vulnerable to DM bias.
Let's say the player wants to spin around in an arc with his sword and attack all the opponents around him. Sounds reasonable enough. But without rules to describe the action/power, it comes down to the DM to decide how the action is resolved. Is the attack made at a -4 penalty? A -2 penalty? No penalty? This can be tough to decide and can have a big impact on the balance of the game. And how often can the player do this? Is there a limit? Why not attack enemies in a circle arc whenever you're making an attack? The overall effectiveness of the class depends heavily on the answers to these questions, and what one DM allows another might not. This could potentially turn players away from the system since the effectiveness of their characters depends on who they're playing with.
Spellcasters don't have this problem since they have clear rules on how their powers work and when they can use them. Their power is a lot less dependent on what DM they're playing with and they don't have to ask for permission when they want to do something cool. They feel in control of their own characters, and that's very important to a lot of people, including myself.
I understand how having to know how to perform a whirlwind might be iffy from a simulationist point of view, it makes for a much more concrete, efficient and reliable game, and I think that's important for such a big product as D&D. That's what I want from D&D Next.