I've dumped a bunch of information on the other Jon. Waiting on him now if he needs anything more.
The brief summary is:

• Most of the notable work with stylesheet content in the head we did was done in May/June and the async change in July we reaped the benefits.
• The graphs on http://grafana.wikimedia.org/#/dashboard/db/readingweb are our most useful tool for showing how the mobile site has improved over quarter.
• First paint is lower on mobile than desktop on the same connection:
  • Special:Blankpage [1,2]: 1.681s vs 1.872s
  • Facebook [3,4]: 2.673s vs 4.384s
  • Barack Obama [5,6] 3.165s vs 4.283s
• Problematically we have little historical data to accurately measure how impactful our changes have been
  • Going forward https://grafana.wikimedia.org/#/dashboard/db/webpagetest will give us a more reliable way of comparing past and present
• We did little tangible towards performance this quarter. The real successes were around uncovering big problems with the mobile site and forging plans to address them during the next 2 quarters (image and HTML size).
• By giving attention to performance we've been able to notice regressions (for example the changes to Echo on desktop impacted our first paint). Getting to a point where we can detect regressions is an important next step.

[1] http://www.webpagetest.org/result/150928_0Z_1BV3/
[2] http://www.webpagetest.org/result/150928_EJ_1BTW/
[3] http://www.webpagetest.org/result/150928_Q5_1BTF/
[4] http://www.webpagetest.org/result/150928_B5_1BTE/
[5] http://www.webpagetest.org/result/150928_YY_1C09/
[6] http://www.webpagetest.org/result/150928_1V_1BZF/

@Jdlrobson - all good thanks!