First of all, having watched his FPVoD, my foremost reaction was empathetic happiness. It's a little sad to me that the reaction to his accomplishment is dominated by discussions of balance and game design. We are fans and viewers of a competition. A massive underdog victory by a player who seems to me to be nothing but a good-willed competitor should be something that sparks joy in us all (except diehard Zest fans but I'd hope even they could spare a moment of exploration from the other perspective and enjoy it as well).
That said, there's nothing wrong with using this event to do a little analysis of the state of the game. But I never see what I consider the most important point taken into consideration: taking control of the game. There is an extremely high value in dictating the path a game takes. Progamers can hope to force their opponent down a certain path by doing certain things that should elicit the hoped-for responses, but it's not an exact science.
Cannon-rushing remains powerful to this day because one player can completely dictate the path the game will take by making a forge and a proxy pylon. Once that is accomplished, the most important factor in determining the winner is all the technical knowledge involved in executing and defending cannon rushes. All other game knowledge and skills become mostly irrelevant.
This can happen for other builds, for normal builds. Mostly we think of aggressive builds doing this, like Parting's Soul Train, or Parting's 4gate blink, or Parting's... well, mostly we think about Parting coming up with an aggressive build and the game hinges on Parting's ability to execute it and his opponent's ability to defend it.
But it can happen for macro builds, too! Think about someone like Stats who safely expands and defends up to 3-4 bases and, at most, sends out a warp prism with 4 zealots as his only aggression before he's nearly maxed out. Such a playstyle leaves the opponent a lot of choices for attempting to crack it, but often the fact remains that only one choice doesn't cede some kind of advantage to Stats: expand and macro too. So it can be quite similar in practice: attempt to interrupt Stats's gameplan in a disadvantaged way, or follow along with his gameplan and get outplayed because he's the best macro and late game player.
Cannon-rushing is the best version of this. It's the absolute best way to dictate the path the game takes and it prunes away the maximum amount of general game knowledge and skills, making victory determined by the smallest set of skills and knowledge possible. Be better than your opponent in that small set of skills and knowledge, and you win.
Is that unfair? Is that so bad? Does it cross the line? The funny thing is that if it actually was too good, the problem would take care of itself. Cannon rushing would become so common that everyone would get better at defending it. Well, pros would. Most players are hardly getting better at all. Any solution that involves players having to learn something new and improve is not helpful to most players. That's why cannon rushing is so frustrating for most players: it's an extremely condensed and clear demonstration of their inability to do better on their 2nd, 3rd, 10th, 100th time facing the exact same thing.
Anyway, back to pros. Right now cannon rushing occupies a weird space where a few dedicated cannon rushers have elevated their cannon rushing game far beyond what any occasional cannon rusher can achieve. Any player who is not privy to what the best cannon rushers can do will absolutely lose games to them, even if they're pros. If the pros got enough practice against it, they'd become virtually immune to it. But they don't, so they aren't. And why don't they hit them on ladder, at least? Because these cannon rushers struggle to hit MMR's that match with pros, dragged down by lower win rates against zerg and terran, and dragged down by players recognizing them and metagaming them for an easier defense. And, in fact, dragged down by amateur players who get so much more practice against cannon rushers that they're better at defending it than some pros are.
I really enjoyed ProbeScout's victory from a game design perspective by respecting the power of dictating the path of the game. I'm not convinced that it's overpowered. I believe some players are far better at doing it than some other players are at defending it, but with equal practice doing it and defending it, the defender has the advantage. So that's fine. I think it'd be most excellent to see a progamer plan to do a cannon rush on a map in a bo5 and actually execute it at the level of a dedicated cannon rusher, since some progamer attempts at cannon rushing in the past have been pretty poor showings.
I personally expect to lose to a lot more cannon rushes but I hope to enjoy the process of improving against them.