Well one thing is for sure.. Its not going away. We have only scratched the surface of the technology's potential applications. The printers themselves are more or less still at the level of old school dot matrix to make a rough analogy.
Unless the future is so bleak we can't innovate like we did in the last century where computers went from filling a room to barely filling the palm of a hand, few people having large heavy video players to tiny units that can be carried in small bag. From big massive printers whose inks were costly to cheap units that can be thrown away and the inks are justa few pounds then I woul guarantee 3D printing will eventually be consumer level, gesture and speech control will be the future until the human brain can manipulate wirelessly. You people really need to get your head out the sand and look at history for a vision of the future.
I think expectations should be moderate: yes, millions of people will buy 3D-printers in the future and no, this own't make factories obsolete, after all, factories can always buy bigger 3D-printers and use them more efficiently than you can at home.
I think the thing most skeptics have a problem with seeing is how a 3D printer helps with anything but shipping cost. Yes, you can 3D print something like a gun, but how many gun buyers want to go through the pain of manufacturing their own gun? You can get almost anything next day shipping if you want. 3D printing will gain traction at some point, but it's got a steep embedded paradigm to overcome. Fans of 3D printing will have to be patient.
3D printing has been around for a long time, aka rapid prototyping. The problems are size limitations, materials, resolution, and speed. Yes, speed. They are very slow to make parts as the polymer needs to be cured. You cannot make very large objects (usually limited to about a cubic foot), and the materials are not the greatest - strength and finish. The surface finish is not very good, even with high resolution printing.
But they are good for the original intent: prototyping. Mass production, not so much. Actually, not at all. In the future, maybe, but that's a long way off. We use the latest and greatest Stratasys equipment and it has a long way to go.
What is clear to me is that this is more of a manufacturing tech as it is a consumer one, and it's for those who like to manufacture something and are willing to educate themselves how to do that. In the same time it will shift the definition of manufacturing quite a bit and will most likely make it more wide and likely will make more people willing to participate in manufacturing of physical objects /and that is not a small thing/.