It's viability mainly comes from the ASIC-resistant nature of the project, and the steps the team have taken to ensure small miners stay relevant.

For example, the uPlexa team already have an emergency fork ready, should anyone develop a compatible ASIC - one which they can deploy in hours, rather than months.

On top of that (and I'm not a developer, so they may chime in here), I believe the algo used is 'kinder' on low powered devices (half iterations from the original Cryptonight iirc).

The main note from the project thus far, is that we're actually seeing delivery of these promises, and when we're surrounded by projects that simply don't deliver, I think that has to be a reason to show an interest