It's probably fairly unknown. I hadn't heard of it until today, although some poking around shows it got some exposure in 2016.

Also, for what question is it exactly the answer? My first impression is:

"I just want to continue to run my Python2 code": it works for that, but I'd probably go for Python 2 as included in e.g. RedHat or Ubuntu LTS versions for now, which will provide security support for a few years more, over a fairly small project with unclear future. Or Google App Engine (which is the main reasons one of the projects I help maintain still supports Python 2)

"I don't want to upgrade to Python3, but still use new features": Tauthon is probably the only solution, but you're now writing software that only works in it. Are you willing to use a fairly niche interpreter for your project, and hope it gets maintained and supported as you need?

"I want to use it to step-by-step migrate to Python3": I'd have to investigate that more, but I'm not sure it provides much more in the way of compatibility than 2.7 does with __future__ flags. In some places there might be an argument that you could get more benefit from new features during that process using Tauthon, but I think you'd still have that final push to switch over at some point.

And for software you ship to customers, depending on the customer set switching to a "weird" (for whatever their IT thinks that is) interpreter isn't really an option.

I'm not a developer so I'm just genuinely curious.

Does the work of migrating to Tauthon (which does sound simple) all that different from just porting to py3 if you actually need those py3 features?