Holy shit, it's been two years and unsurprisingly, I was somewhat wrong about the decline in motion controls. Contrary to where I thought virtual reality was going, much of it has tried its damnedest to reestablish motion controls in gaming as a means to embody, or "immerse", players in virtual reality. As virtual reality's staunchest advocates will likely tell you, you haven't experienced true virtual reality and "presence", or as I'd be inclined to call it, "embodiment", until you've tried it in room-scale with motion controls and the ability to walk about the virtual environment.
Despite this, and as numerous, better written articles have noted, this only presents new problems to long unresolved existing problems with motion controls. I would recommend researching, albeit with a warning if you have been victim yourself to these experiences, virtual reality violation and harassment, to get a sense of the problems I'm considering in the previous sentence.
Enduring Motion Control Problems
That said, let's cut to what I had in mind a couple years ago when I had meant to address just why many developers had decided against adapting to, or failed to adapt to, motion controls. This will largely be a higher level discussion, with minimal attention to the lower level details of how these technically work, but with a nod to them where I believe my understanding may not be terribly off.
As noted before, the Wii was a massive success, and yet few of the major developers appear to have really tried to build upon it. While Nintendo itself has a mixed history with third-party developers, the overwhelming proliferation of "shovelware" titles on the system suggests that this may not have been the major problem. Instead, it may be attributable to the technology employed in the motion control system being much too constrained and inaccurate to permit the finer controls developers may have desired.
This is entirely possible, but then, as several titles demonstrated, you didn't necessarily need fine control for swinging swords around (e.g. Red Steel 2, Legend of Zelda: Skyward Sword) and in fact, where precision was a concern, such as aiming, it actually tended to work rather well (e.g. Metroid Prime 3, Red Steel 2, Conduit 1 & 2).
Although, in each of these cases, were one to get too excited, tracking could rapidly be lost and make for brief, inconvenient moments of lost interaction, varying depending on how the developer handled this (e.g. pause-prompt to continue disruption to avoid unfairness). However, generally tracking could rapidly be restored, so this doesn't necessarily seem like it could be marked as a major knock, but it would greatly depend on the developer's handling of it.
Missing Creativity and Key Lessons
Despite these real obstacles to smooth gameplay that may emerge with motion controls dependent on the state of the technology or its general configuration, what seems to emerge more than anything is simply absent creativity. Lacking finer controls, such as easy methods to wiggle each individual finger, or to make the hand or appendage in the virtual space physically interact with the virtual environment, developers seem at a loss for what to do with three dimensional cursors.
Which is, essentially, what all these hands and limbs manipulated with motion controls are. This isn't lost on developers, of course, except that many, uncertain of what to do with the depth aspect of this interaction, often turn to limiting it to literally acting as a two dimensional cursor. To be fair, depending on the type of motion controls being discussed here, some are less capable than others to read depth at all, meaning that this might only be an exercise in frustration to even implement with them.
However, with those that are capable, one need only think of what they can do with their own limbs in life and what silly antics they can get up to when toying with their own reach to develop fun gameplay.
Even with those less capable of reading the depth movements of their controllers, if the limits of the tracking can be taken well into account and worked with, rather than against, one can uncover a number of captivating gameplay ideas that would be much too awkward and cumbersome with analog sticks alone. Games such as Okami should have been a natural fit, but whether it was a bad port or the technology's constraints, it revealed that such a game would need to be more carefully refitted for motion controls.
Red Steel 2 is an excellent example of a game that recognized its missteps in the first, and improved significantly as a result. Unfortunately, more developers appear not to have even made the first attempt with which to learn, and even now as they are, some seem to be neglecting the experiments of their predecessors on the Wii and the lessons that were learned, much to their disadvantage.
That said, the degree to which these lessons learned have been thoroughly shared among developers is something of a mystery to me. Considering the degree to which developers tend to keep certain techniques up their sleeves, whether out of contractual agreements or otherwise, it would be little surprise to find that few of the techniques learned have been shared, which is incredibly unfortunate.
Then again, that's my ignorance and my cynicism speaking. Hopefully I'm yet again wrong in this respect, but if I am, it makes it all the sadder that the state of motion controls remains so poorly explored.