Again, the reason I posed the question about roulettes, and how much systems should condition their players towards particular choices in the duration and locations of their playtime, is that as the game is increasingly streamlined, efficiency increasingly takes priority over engagement. We see this in the developers' designs and that does trickle down into player and community perceptions as to the shape and therefore intent of the game. (They then make requests for efficiency over engagement in ways that are easier for the devs to meet than requests for engagement over efficiency, and the feedback loop spirals on.)
I like that they exist. I just feel that given their current efficiency bonuses, they take too large a part in the game. Or, if you want to look from the surrounding content inward,
I feel that too many other systems have seen their rewards stagnate as to be made increasingly less a part of the seemingly intended experience, which then narrows the game -- excessively, in this case.
I feel that excessive narrowing of the game harms the experience. While some structure is certainly a boon, and in catching up there are some parts that must be dispensed with to focus on a more core shared experience between players, when the shape of any game centers seemingly on efficiency or just the tools to accomplish something, it makes the means seem that much less important. Sure, players can go against the grain and ignore the signs and try their best to perceive things purely on their own criteria, but most experience games largely through their shape and apparent intent.
Now, is that all centered on Roulettes giving X bonus tomes? No. But I think we need to be less eager to ignore the fine details in favor of stark changes alone. Most losses veteran players will have noted to their experience over the years, especially if playing more than a single or few jobs, can better be attributed to the dozens of cuts of indirect changes than any one direct change.