The memory issue the post mentioned was a really long time ago, I remember they did Generational GC, Compact GC and the most important project from Mozilla, MemShrink from Nick. That was before e10s and all of that was what made Firefox memory efficient today. ( I think Nick is working at Apple now )
The failure of FirefoxOS, was mostly because of two opposing goal or ideal. They want the OS to be web based and everything to be made with JS or Web Technology. The believe of JS will be good enough and that JS VM will be fast enough runs deep inside Mozilla at the time to be point I remember FirefoxOS engineers were even afraid to speak out on why dont they just code this in C++. The other goal was they want Firefox OS to be running on low cost $50 Smartphone and somehow thinks the Smartphone hardware price will drop while performance continue to improve. I was surprised, and it was only many years later I realise most software developers have absolutely zero understanding of hardware, BOM cost, supply chain, and long term component cost development. And I still think, FirefoxOS was a distraction for Mozilla. Spreading themselves too thin while battling Chrome on PC. They really should have fought Chrome first before their move on Smartphone. And 10 years later we really do need a third OS option in the Smartphone market. Although arguably Firefox OS still lives on as KaiOS.
But again, I say this a lot on HN, and as someone who pushed through hundreds of installation of Firefox in different places, nearly every single response of switching away to Chrome was because Chrome was faster. Way faster.
And I thought it is a good time to remind ourselves, good faith and ideals only last so long. Ultimately you need to have a better product to retain your users.
Edit: And there was another point missing from the article. Mozilla refuse to support DRM and H.264 codec, a lot of users simply wanted their browser to work just left.