He's trying to invent a problem to fix, and then to add insult to injury, he's fixing it wrong.
Start by defining the problem: rolling too many d6s is not everybody's cup of tea. It might be some people's cup of tea, but we want D&D to work for everybody.
How does giving everybody sneak attack solve this problem? It doesn't. It does build a framework in which to think of sneak attacks a little bit differently, but we don't need that framework once we understand what it is we want to do.
Giving sneak attack to everybody is a solution to a different problem, one which I'd be happy to see addressed, but please, in the right context. Sneak attack for everybody is a problem of verisimilitude in a simulationist model. Which is a fancy way of saying "I want the game to be more realistic".
What we haven't done is solve the "too many d6s" problem. If we reframe sneak attack as an optional thing that you can scale, then it will become obvious to any player that there are diminishing returns with choosing more and more and more d6s. The end result is that there is an illusion of choice "Gee, I can take 'extra damage' every level and ultimately have 20d6 sneak attack: Leeeeerroooy Jjjjjjjjjjjjeng-kinssss!" But in reality no player will ever take all 20; instead they will opt for the other abilities.
And speaking of "other cool abilities" a little alarm always goes off in my head when a designer says "we're going to insert something cool here". When you try too hard to be "cool" you often fall flat, and this happened way too many times with new powers made for 4E.