Interesting last comment (from November 2016) to the original article:
"Important Update *
I am currently employed at Pluralsight, and I feel obligated to point out that the author of this article, our CEO, no longer believes or stands behind any of the content in this article. Pluralsight does not operate under the Deming philosophy. As of January 2017, Pluralsight will be implementing sales commissions again. We have reverted to the typical high-pressure, high-stress quota models found in most companies.
I'm sharing this because I don't want anybody to read this article and follow the bad advice found here. Maybe commissionless sales organizations can work, but apparently not at Pluralsight."
You might also want to look at ren-C (https://github.com/metaeducation/ren-c), a fork of the open sourced Rebol 3 (which isn't an active repo). It can be build with an embedded TCC so you can write C in-line for functions where you need a speed boost.
In the older post you mentioned one of the performance issues being that JS couldn't write directly to the GPU for security reasons and so a copy to the GPU would have to be done after the browser vetted things.
To my _very_ naive thinking, and if video RAM was plentiful, couldn't things be set up such that the JS runtime _could_ write to the GPU, but to a reserved section of video memory that wasn't directly executable? After being checked by the JS runtime, bytecode could be copied over to another area in video memory to be run. Would something like this work? If so, has this approach been tried by anyone?
Thanks so much for the upgrade!