I can see most people misunderstood the poster in the thread you linked. The thing you quoted is probably a result of the same misunderstanding, the net effect of the change me and the other guy are proposing is not to make the game any harder or easier, just more logical and streamlined. There are states where you've lost any chance of making it but it only becomes apparent some time later. That's kinda like dead ends in old adventure games (albeit lesser in scale).
I don't know anything about implementation difficulties, that's not my concern as someone suggesting a change (how can I know?) so all I'm looking forward to is people actually getting what the post is about and answering my question: why is a continuous drain better than a quantified one, i.e. one that links directly to your 300s 100s 50s combos, misses etc.? I'll bet you could automatically generate the quantic equivalent drain for each beatmap's current continuous rate by using a relatively simple algorithm to establish how difficult the score is to pass (i.e. what kind of error threshold/margin it has (as a whole or even per section if that's part of the beatmap architecture) not how many notes there are or how fast they follow each other or any other such parameters).
So once again, you still have drain but it's clearer if you're alive or dead, which produces the psychological effect of the game being fairer.
Feel free to merge this with the other thread BTW if someone can do that since I'm saying pretty much the same thing here.
