easiest and sure-shot way for a 2500 GM to become a 2800 or even 3000 rated GM

This theory just reveals a total ignorance of how the elo system works.  The difference in elo between two players will basically tell you statistically how often they should win/lose/draw against eachother and the difference in the points gained/lost compensates for this.  So if a player of 2500 strength only plays against 2100 players, their rating will stay the same unless their playing stength improves.  

This theory reminds me of the martingale betting system where you double your money every time you lose until you eventually win your money back with interest.  It seems to work because you slowly but surely gain money, but statistically you are guranteed to lose all of your gains eventually.  In this system the 2500 player would gradually gain ratings points, but when they eventually lose/draw all their efforts will be in vain.

This theory just reveals a total ignorance of how the elo system works.  The difference in elo between two players will basically tell you statistically how often they should win/lose/draw against eachother and the difference in the points gained/lost compensates for this.  So if a player of 2500 strength only plays against 2100 players, their rating will stay the same unless their playing stength improves.  

This theory reminds me of the martingale betting system where you double your money every time you lose until you eventually win your money back with interest.  It seems to work because you slowly but surely gain money, but statistically you are guranteed to lose all of your gains eventually.  In this system the 2500 player would gradually gain ratings points, but when they eventually lose/draw all their efforts will be in vain.

A "minor" flaw in your system is that the 2500 player has a really big chance to lose and draw some of the hundreds games to the 400 points lower rated players, even if it would be all 2100 players. In addition to that he will have to non stop beat 2600 rated players to get to 3000 (and maintain the rating diff of 400).

My estimation is that in the long run this system will result a rating of about 2500....

This.

The crucial point is although you don't draw or lose often, when it happens it is a catastrophe. One draw at 400 points of difference will cancel out about 100 wins, rating-speaking (from memory, not sure about hte exact number).

How do we know Obama was born in America?  Because he passed a rigorous background check during the vetting process that confirmed this particular qualification of his.

You're also overlooking the fact that as his rating increases, the amount he gains decreases and the amount he will lose when he loses a game will increase.

Is this guy a troll? I think there is a major flaw here that no one in mentioning. It is the getting better part that is difficult. Let’s say he does always play people who are rated four hundred points below him. That means this player has to be good enough to beat those people. That isn’t so bad when you’re a 1200 rated player but what about when you’re a 1900 rated, or 2500. You will have to be able to beat these tough players and since when is that easy? Again not to mention the larger the gap between you and another player is the less points you gain so you could no keep working this system indefinitely. Eventually your points will converge on a value.

Yes but the difference between the 2500 player and his opponents will increase as his rating increases.  Unless you're suggesting that the player who started at 2500 always plays against players 400 points below him even as his rating improves.  In which case it will become increasingly difficult for him to maintain his winning percentage unless his actual playing strength improves.

Also when you say things like, "since when has it become difficult for 2500 or 2600 to beat or draw 2000 or below?" you're missing the point.  Statistically a 2500 will win/lose/draw a certain number of games against a 2000.  Obviously the winning ratio will be high, but the elo points awarded for a win and/or deducted for a loss are designed so that, as long as the players results reflect their ratings their rating will not change.  What you're effectively suggesting is that it is possible for players to play above their rating by playing weaker players.

You can't game the system, at least not in the way you are suggesting.

anybody help me..  Is there any fide rated open tournament in chess.com... 

the starting point of this thought was the fact that im currently in a tournament (OTB) where a GM has participated...the next rated player is 2240 or so...the prize money and atleast 6 to 7 point increase in his rating are already his...i just had a wild thought if he cud do this month after month, playing in such tournaments...all d prize money and 6 to 8 points are assured for him...i dont think he is going to even draw a single game in this tournament out of 9...if a 2600 GM plays a bunch of 1800 to 2000 players consistently, what are the chances he will actually lose one?

Maybe in 9 games it is possible he does not draw a single one. In 1000 games, no chance.

This theory just reveals a total ignorance of how the elo system works.  The difference in elo between two players will basically tell you statistically how often they should win/lose/draw against eachother and the difference in the points gained/lost compensates for this.  So if a player of 2500 strength only plays against 2100 players, their rating will stay the same unless their playing stength improves.  

This theory reminds me of the martingale betting system where you double your money every time you lose until you eventually win your money back with interest.  It seems to work because you slowly but surely gain money, but statistically you are guranteed to lose all of your gains eventually.  In this system the 2500 player would gradually gain ratings points, but when they eventually lose/draw all their efforts will be in vain.