Jeff has made it clear that Discourse lacks the resources to dedicate to this problem. The chrome team has made it clear that this is a tough problem with no easy fix, and they have other priorities. Stef's efforts are heroic but he already has three jobs (Yahoo, core-team, ember-cli).
It appears that Ember in particular, and JS frameworks in general, are in danger of losing their major advantage over native apps. Why not use UIKit if iOS is the only viable mobile JS platform and you have to make a native android app anyway? If you have to support mobile then the only place for JS frameworks is on the desktop.
The Ember community may not have the expertise as a whole to fix chrome on android, but we do have other resources. Can we throw money at the problem? I would pitch in. Bustle puts their money into fastboot; maybe Yahoo, Intercom, Apple, Yapp Labs, Zendesk, LivingSocial, Square, Addepar, Twitch, and Groupon could spare some time or money to form a team to address this issue.
The Discourse meta thread on this topic has ended in resignation that we must wait for newer devices, but a Bountysource might make Ember relevant on mobile prior to 2020. I think we should consider trying to pool our money to make that happen. Other suggestions welcome.