This looks interesting, but I'm not well versed at evaluating a new web framework. I've built a few projects in Django, and I have a basic understanding of how I'd go about building a project in Flask. I've never worked on a high-traffic project.
Feature-wise it seems close to Flask. I suspect that Flask is more opinionated than Falcon: it requires the Jinja2 template library and your project may end up with a tight coupling with Flask if you use its blueprint system or don't think through how you use Flask's context variables flask.request and flask.g (Flask tries to be too clever here IMO).
Falcon looks more like a library than a framework to me and its design seems less likely to affect how you structure projects compared to Django in particular but also Flask.
Usually, when you are coding a full API in Django or Flask you find yourself putting additional dependencies in, and this will impact performance and maintainability. Those dependencies won't help you get rid of the "opinion" or "initial purpose" of the frameworks: to build websites, and this will also impact maintainability and performance.
If you are at any point returning HTML from your Falcon project or thinking of serving static content, you should probably not be using Falcon at all. :)
Django/Flask are written for the programmer. I think Falcon is, to a degree, written for the machine (although if you're a systems programmer you may snark at this).