EDIT: I read further and it seems like he enjoys poker. So I guess it's possible that he'd enjoy MTG.
To me, this is more "fun". With chess, it's just brutal. Unless both players are evenly matched, it's a steamrolling. No fun for either player. At the top tiers though, I can see it being very interesting. But then again, I find watching top players play Hearthstone to also be very interesting. Especially when you consider the meta-strategizing of picking certain cards for their decks.
You just have to play a handful of games in order to get a rating (whether in real-life tournaments or online), and then you know that playing somebody with a rating within 100 points of your own rating will usually result in a good game where both sides have chances to win and neither will consistently steamroll the other.
The only times that happens as far as I know are when the rating really doesn't reflect the current skill, which can happen over the board with juniors that are improving so quickly that their rating hasn't caught up yet, or people who haven't yet played enough games for the rating to be accurate (e.g., a brand new account on a site that starts with a rating of 1200, which is solved by not playing people who haven't played enough games yet).
The article discusses how different players prefer different ratios of luck vs skill, so making the most "fun" game in this regard really depends on who your target audience is. Similar to your thoughts on randomness. :)
I think that makes a game more fun to participate in but significantly less fun to watch at a competitive level.
I am someone who enjoys poker and watching chess on TV (usually an internet broadcast). The problem with chess as a spectator event is that, at least for me to enjoy it, it takes a lot of time and effort, and there is no social aspect. For example, I enjoy analyzing a game of chess myself, but it takes hours of my time and also computer analysis, and then eventually I get many "aha!" moments, once I understand why a move was (or was not) played. That's really hard to translate to a general viewing audience.
Compare that to athletic events. Anyone can watch and marvel at those who are bigger, stronger, and faster because those properties are well understood. Even without any knowledge of the sport, you can watch a sporting event with your friends while having a beer and have a great time. Probably not going to happen with chess.
With poker, at least you can cherry pick the improbable outcomes and daring bluffs, and add drama due to the high stakes often involved. It makes for very suspenseful watching when a guy risks more than our annual income with a junk hand, and he pulls it off.
So my memory is failing me and I was off by a decent amount, seems mid 60's is an average ceiling for win rate.
This is just an example from the 2013 season at the world championships. You'll want to sort it by Con(structed) Win %. Even the best players in the world win slightly more than they lose over an entire season.
EDIT: Actually individual hall of famers might be better to look at, especially since these are lifetime stats.
http://archive.wizards.com/Magic/magazine/halloffame.aspx?x=...
http://archive.wizards.com/Magic/magazine/halloffame.aspx?x=...
If yo go to the career stats page you'll see that kibler has a lifetime win rate of around 60% and finkel is slightly higher than that.
This may also hold for chess.
The point here is obviously that Magic has an element of chance that chess lacks.
You can play theoretically superior chess and calculate more positions than your opponent and still lose the game, although it's very unlikely. Both player have a horizon and sometimes they can bump in a position of clear superiority for one side that none predicted.
If such a game could be foiled by the other player making a "bad" move, then it would by definition not be a bad move.