Intertnet surveys are unfortunately not a good source of unbiased data.

You have a better idea?

 Yes, a couple. I'm working on one of them right now, as I stated earlier in this thread.

The most difficult problem to account for when attempting to convert a rating awarded by one chess organization into an 'equivalent' rating awarded by another is the fact that most,if not all, of these rating systems are (over the course of time)  in a continuous process of modification due to administrative tinkering.

The typical chess organization adopts a reasonable formula for the calculation of ratings, (Elo ,Glicko,etc.) and is then unable to resist the urge to modify it.

More often than not, this is done with the best of intentions. Nevertheless, the integrity of the formula itself is inevitably compromised.

Yes that's certainly true, given a few variables to tinker with people cannot resist eg contemplating increasing rds (already too high at chess.com) for no good reason other than you can.

Overall for me the various governing bodies and also the new dot com outfits are doing a good job in aggregating the grades.

Anomalies will always happen on the net for personal reasons, centaurs etc. But its hoped these are a minority.

>:)

In England, we use BCF ratings which can be converted into a FIDE rating ......

Only the ICC "autopairing" is a good reflection of OTB FIDE ratings, as players cannot choose their opponents. There is a lot of rating inflation on the ordinary 5 minute ratings on the ICC (e.g. many GM's over 3000), but far less on the 5 minute autopairing (max about 2700). 

Within this, the kind of titled opposition (FM's, IM's, GM's), and the fast time control is a good simulation for playing in strong OTB events, where intuition, speed of calculation, knowledge of openings, and good endgame technique means that you should get a good FIDE rating, if you can do well in tough blitz environments like the ICC autopairing or the Chessbase server where you don't try and inflate your rating by choosing people with low risk of losing.

Correspondence chess ratings are completely different and they are more of a reflection of accuracy and concrete deep analysis. Often many titled players have very high ratings on the major blitz servers, so I think blitz chess has a higher correlation because it lets the stronger players demonstrate more clearly their faster intuition and greater opening knowledge, etc, than correspondence style allows.  

Really?

Germany used to have their own rating system too, called ingo zahl or something like that. I wonder if they still do ? I think it would simplify things if all countries went by fide/elo ratings myself. I had an ingo-zahl when I was in Germany (80s) of 96 or 97 , which I didnt understand what this meant. I saw GM Larry Christiansen playing in a tourney while there and his was like 39, which is the lowest I had seen. But the lower the number the stronger the player !  I dont know if they still use this rating system or not but I never did get used to it.

There are a lot of different rating systems.

I guess a lot of players like to play blitz games on the internet.

Thought the blitz ratings cannot accurately be compared with ELO or BCF ratings I have found a table of comparison which might give you an indication of what your position could be in another rating system.