I'm agree with RealityMate. A draw in top level playing is a great achievement and some of the best games ended as a draw, so it deserves to be awarded properly. I think that the best measure is to disallow draws by agreement. In chess there're ways enough for players manage to have a draw game. Let them take the risk of playing until perpetual checks, threefold repetitions, stalemates, etc. It's enough to increase chess appeal...
5-2-0 point system

I totally understand your point. However, that's exactly why I think the 5-2-0 is good: unlike the 3-1-0 system, it doesn't force people to play like Tal, but the difference is enough to discourage draws on move 20. In reality, it's almost exactly the same as the standard 1-.5-0 system. 5:2:0 = 1:.4:0
I don't expect either this or 3-1-0 to ever come close to replacing the standard system, but I think it would be an interesting alternative to have in some tournaments.

Just out of curiosity, I tried converting some standings from recent tournaments to 5-2-0 format.
Big thank you to SonofPearl for your newsposts!
Name | Elo | W | D | L | Pts |
Kramnik, Vladimir | 2780 | 2 | 4 | 0 | 18 |
Anand, Viswanathan | 2800 | 1 | 5 | 0 | 15 |
Carlsen, Magnus | 2826 | 1 | 3 | 2 | 11 |
Shirov, Alexei | 2749 | 0 | 4 | 2 | 8 |
W | D | L | Pts | |
Alexei Shirov | 3 | 2 | 0 | 19 |
Levon Aronian | 1 | 3 | 1 | 11 |
Vladimir Kramnik | 1 | 3 | 1 | 11 |
Wang Hao | 0 | 2 | 3 | 4 |
Name | Nat | Elo | W | D | L | Score | |
1 | Kamsky, Gata | USA | 2713 | 9 | 2 | 0 | 49 |
2 | Gashimov, Vugar | AZE | 2719 | 8 | 3 | 0 | 46 |
3 | Aronian, Levon | ARM | 2783 | 9 | 1 | 1 | 47 |
4 | Bareev, Evgeny | RUS | 2663 | 9 | 1 | 1 | 47 |
5 | Karjakin, Sergey | RUS | 2747 | 8 | 2 | 1 | 44 |
6 | Kempinski, Robert | POL | 2612 | 9 | 0 | 2 | 45 |
7 | Grischuk, Alexander | RUS | 2760 | 8 | 2 | 1 | 44 |
8 | Shirov, Alexei | ESP | 2749 | 8 | 2 | 1 | 44 |
9 | Naiditsch, Arkadij | GER | 2684 | 9 | 0 | 2 | 45 |

I totally understand your point. However, that's exactly why I think the 5-2-0 is good: unlike the 3-1-0 system, it doesn't force people to play like Tal, but the difference is enough to discourage draws on move 20. In reality, it's almost exactly the same as the standard 1-.5-0 system. 5:2:0 = 1:.4:0
I don't expected either this or 3-1-0 to ever come close to replacing the standard system, but I think it would be an interesting alternative to have in some tournaments.
Well, I don't think 1:.4:0 is almost exactly the same, because 2x .4 doesn't make exactly 1 :). Even .499 doesn't quite get there...
If people want fewer draws, you need a substantial rule change, maybe like stalemates are not draws, but wins. (it used to be considered an inferior win, long time ago.)
But in a game where a draw is a valid result, playing with the points value just seem wrong.

3-1-0 a draw is almost as bad as a loss, I mean really. If I had been losing and suddenly I can fight for a draw, at the end of the game I'd feel crappy to get the draw... I would want at least one of us to get 3 points, with a draw it's like we both lost
5-2-0 isn't so unreasonable. An occasional top tourney with this wouldn't be bad. But this is just reminding me how I hate how some things have changed (I'm not even old enough to have enjoyed the good times) e.g. proper candidates cycles, longer WC matches, (what the hell is a blitz play off?) multiple time controls with adjourned games, and none of this 3-1-0 stuff.
I like the 2-1-0 ratio best, and hope it's never replaced. Draws are a natural part of chess, and less than 1/2 of a win seems to be punishing them. On top of this, it's not as if FIDE or in my area the USCF has exhausted all other avenues of sponsorship and publicity.
When reading the poll about 3-1-0 scoring system used in Bilbao, and tossed out the idea of a 5-2-0 system, which would discourage early draws, but not as drastically as the 3-1-0 does. A couple people liked the thought, and it didn't seem to have any obvious flaws, so I figured it would be worth posting in the forums where people could argue its pros and cons.