In Version 1.2, there are options of molecular dynamics algorithms (see attached image) for geometry optimization, which is useful in some cases. Maybe include the function in the coming versions?
It is on the roadmap. 
I believe Geoff’s primary goal/condition for 2.0 is that it should essentially have feature parity with 1.2, so if you take a look at the roadmap and find that something key to your usage of 1.2 is not there (and is not implemented already), I would recommend bringing it up!
Yeah, the bigger priority is clearly implementing UFF internally and making sure all that code is fully crash-proof.
Adding basic MD will be easier once that’s done.