I agree, open model sharing is a great benefit plus it empowers people in using their own creativity.
This is a democratization of traditional plant-based manufacturing.
My vision is that in the end, SLS or hybrid systems (for example for multimaterial) are the way to go once they are more automated and prices come down by say, five times. I think we will be there in 10 years. The reason for SLS is speed and homogeneous visual and mechanical quality.
Making 3D printable into a major global manufacturing technology will require overcoming some disadvantages.

- Heterogeneity - layered product buildups leads to major failure for mechanical loads, often clean splits along the weakest layer. We can overcome this by printing truly in 3D i.e. the printhead moving also in Z-direction to create waved or zigzagged layer structures. That is why I think Scara systems will be a good approach. For multiple materials and different levels of detail, we will simply have multiple robot arms each with their own extruder working on the product.

- Production speed - Lulzbot is handling this pretty well with the moarstruder and automatic bed leveling further automating the process.

- Error handling, maintenance - 3D printers simply need more sensor integration for dynamic feedback to regulate the print and communicate to the user - integrated live webcams are only a start. For example, with the right software a printer can know when the filament will break, when the extruder is jammed, even what the final print should look like visually, report back errors based on image recognition and suggest new settings to improve printing.