Hi U_K!
Challenging ratings are tough to get right. I hope fourth edition will be superior, but I'm not holding my breath and may very well stick with 3.X. Only time will tell. I'm certainly not going to proceed with the assumption that 4th ed will necessarily fix everything. But hopefully you will U_K
!
I love what you've done with challenge ratings. I may nitpick some of the things you've done, but even if I disagree with some points, certainly I commend you on doing something that no one else has done, and making some significant improvments on how CR works for higher level characters.
I guess my main question still though is that I still don't understand why you've done what you've done recently with your ECL CR split and CR equals 2/3 of ECL. I think it's a mistake. I think that ECL should not be used for opponents at all--only used for party members. I think that CRs should be used for monsters, that that CR X should represent, "this creature would be an equal challenge to a well-balanced party of adventures of level X as an NPC of level X with wealth of a PC of level X would, and this creature would be a moderate(easy) challenge to a well-balanced party of 4 level X PCs" [I'll call this tcwbaectawbpofolxaanolxwwoapflxwatwbamectawbpo4l4x with X=20]. This is basically your definition of ECL now, but I think it's better off if you make that definition of CR, because that's basically how everybody else defines CR.
Basically, when the official rules say something is CR 20, they mean tcwbaectawbpofolxaanolxwwoapflxwatwbamectawbpo4l4x with X=20. But when you say something is CR 20/ECL 30, you mean tcwbaectawbpofolxaanolxwwoapflxwatwbamectawbpo4l4x with X=30.
I do think you have an excellent point when you say that PC classes are not necessarily balanced, and a discrepancy of 40% is not that suprising either, especially at higher levels, where the discrepancy probably gets even more pronounced. For playtesting purposes, if you want to determine whether a given monster is an equal challenge to a level X NPC, I think that a psychic warrior (XPH rules) is the ideal test point. So, let's say that I want to propose that a given monster would be about an equal challenge as a level 22 NPC with PC wealth. So, to playtest this, I would put a playtest group of one fighter, one rogue, one wizard, and one cleric against the monster. Then put the playtest group up against a level 22 psychic warrior NPC with level 22 PC wealth. If it's about an equal challenge as the monster, then we can conclude that the monster is about an equal challenge as a level 22 NPC with PC wealth, which would equate to 22 ECL in your system, or a 22 CR by the official rules.
Why psychic warrior? Well, some classes aren't good for testing purposes versus a party, because they're support classes, not really meant for soloing. A class that has options, and which we could consider as a reasonable possibility for solo fighting, would be ideal (for purposes of playtesting CR). The main classes that come to mind for this are cleric, druid, and psychic warrior. But we also want a class that is neither overpowered or underpowered, which would skew the results. Many consider the cleric and druid to be overpowered. Psychic Warrior I see as a good balance as being neither underpowered (e.g. fighter, bard) or overpowered (e.g. cleric, druid).
But that was a digression from the main topic. I still don't understand the point of CR=ECL*2/3. In explaining your reasons for doing this, you said, "This discrepancy is easily illustrated when you contrast a 20th-level PC with a CR 20 monster like a Balor. The Balor is much tougher." If that's really true, wouldn't it make more sense just to say that the Balor's supposed CR of 20 is wrong, rather than to say that CR=ECL*2/3.
Because now you changed the definition of what CR means. To everybody else, CR means, "this creature would be an equal challenge to a well-balanced party of adventures of level X as an NPC of level X with wealth of a PC of level X would, and this creature would be a moderate(easy) challenge to a well-balanced party of 4 level X PCs". But instead, you've given that as the definition of ECL, and have defined CR to be 2/3 of ECL.
And your EL system compensates for that. Because your table now reads, in v6, with CR=(2/3)*ECL:
• Challenge Rating is 1/2 Average Party Level = Encounter Level -3 = Easy
• Challenge Rating is 2/3 Average Party Level = Encounter Level +/-0 = Moderate
• Challenge Rating is x1 Average Party Level = Encounter Level +3 = Tough
• Challenge Rating is x1.5 Average Party Level = Encounter Level +6 = Very Tough
• Challenge Rating is x2 Average Party Level = Encounter Level +9 = Impossible
That's again the v6 version, where CR=(2/3)*ECL. That's a change from the old system, in the official system and your v5, which used to have a baseline of: Challenge Rating is x1 Average Party Level = Encounter Level +/-0 = Moderate. But you only had to make that shift because you multiplied by CR by 2/3.
So it looks like you multiplied by 2/3 to match your CR closer to the official CRs. But then you changed the definition of CR!
So here's what you've basically done. You've said. Hmm, my system, before the change, tells me that Balor is CR 30, i.e., about an equal challenge as a level 30 character with PC wealth==ECL 30. But the offical rules say the Balor is CR 20, and everybody wants to know why my CR is so different from the official rules. I know!! Let's define CR=(2/3)*ECL! Now a Balor has a CR of 20! Everything's the same! But now you have to shift the EL table. And now you've changed the definition of CR when you didn't have to. You should just keep CR the same, keep the EL table the same, and correct the CRs, if they need correcting.
And if you think that a tough encounter is a better standard as a typical enounter, you don't have to re-define CR; you can instead just state that the DM should aim for "tough" encounter (EL+3 or whatever that turns out to be) rather than aiming for a "moderate" encounter of EL +/- 0.
A second question is, how thoroughly have you playtested CR*2=EL+6? I would submit that, at CR 25 or higher, you might find that something like CR*2=EL+8, or even a little more than 8 possibly, is a little closer. You already had a "glaring error" in your initial assumption that CR*2=EL+4 was wrong. But could you be making another error in assuming that CR*2=EL+6 is necessarily right?
Again, I'm nitpicking what I think may be errors in your current systems, but props on what you've done so far on challenging challenge ratings (not to mention everything else)!
) in that the Balor would be a moderate (easy) encounter in v6 to the level 30 party, not tough. But I think that a Balor is not as tough a challenge as a level 30 character with PC wealth would be. 
