I am in the view that the ELO system, developped basically for chess, a two-player deterministic game is not adequate to rank Subterfuge players.

First of all, the ELO system is based on the asumption that a player with a higher ranking plays better than a player with a lower ranking in 90% of the games. This is generally true for chess, but not for our beloved Subterfuge. You may have a higher ELO ranking if you had a good network of alliances, a noob player building a mine next to your 150-driller strong garrison or a perfect starting spot.

Secondly, the ELO system distorts the way you play. At higher rankings, you may be reluctant to join games with lower-ranked players and during the game itself its your best interest to gang up with similarly ranked players against higher ranked ones.

Anonymous games somehow offset those shortages though.

At the highest level, its the best interest not to join games anymore and wait for the downfall of the ones above you.

Of course, all this is valid if you actually care about the Leaderboard. As for me, I try not to build my decisions on the Elo ranking and focus more on getting as much fun as I can from the game.

I think the rating s are fair. I think that to get to a higher level, you must've done something right. Even if it was just luck, you still got there. So having to do very well again to gain rating makes sense. Oppositely, a lower player may not be as good, once again even if it happened by luck, they still dropped. Therefore they don't need to get as high of a finish to move up.

However, that being said, I stay away from ranked. I played them for a while, but they just aren't what I like. I had a friend drag me into one, but normally I'm a regular play type guy.

While I have met a few high ranked players who seemed to be abysmal at any form of strategy or diplomacy, I don't think it is a common issue. Luck can distort a player's rating, allowing an unskilled player to gain 100 points by chance, but in the long run it shouldn't be a problem. Their next rated game should knock their rating down to its correct level. The system is not used to determine someones ability to strategize, it's purely a variable indicator of someones ability to place higher than others. Because of the somewhat random nature of a game based on diplomacy, it will never be 100% accurate, but that is to be expected. Even if the only reason your rating is exceptionally high is because you've been able to create strong alliances to win, the rating is not inaccurate. A higher rating doesn't mean someone is better than you strategically or diplomatically, it just means that according to their previous history, they are statistically more likely to win. Depending on the outcome of each game, that statistical likelihood will be recalculated, resulting in a more accurate system.

Many people are overratted or underrated in chess as well because of luck.

Ok, lets continue with the current system then as I concede that I also believe that distortions will be offset in the long term.
Would just like to remind John Manyard Keynes words: "In the long run, we are all dead"