I have been slugging away at my system for figuring out how monster difficulty should be rated in Heroes By Trade. Doing so has given me more insight into why 4th edition DnD did monsters the way it did. In 4th ed. monsters were all given a level that corresponded with player levels. 12th level players should fight 12th level monsters, for example. Also monsters were categorized by type - minion, regular, elite, or solo. In theory this makes for a very tight system for creating encounters where monsters always have stats that are appropriate for the players. In practice it makes the game encounters feel very artificial because random brigands you meet at level 20 are so drastically more powerful than random brigands you meet at level 1. The world isn't populated with interesting things at that point; instead everything the players fight just happens to have the perfect stat spread.
The thing is that the system described above is just so convenient. I was trying to assign static numbers to monsters that would work across a variety of player levels and it was a nightmare. Monsters that worked fine as part of a level 25 encounter were utterly lethal when faced as a solo encounter at level 10 because the players just couldn't hit them or couldn't get past their armour. There simply wasn't a single number I could assign to them that adequately reflected their difficulty. This is the genius of the 4th ed. model - you have two dimensions to the difficulty (level and minion-solo) and that makes it so easy to figure out because you always know what level the players are.
The major challenge involves the solo monster problem. That is, when one side in a fight has a single HP pool it is at a massive advantage. The other side cannot focus fire to remove threats and focus fire is the primary strategy of any successful group. The flip side of the coin is that a single monster often cannot make use of its debuffs and is far more vulnerable to debuffs than a group. This means that monsters that have debuffs as a primary focus can be more deadly in groups and monsters that have big stats as a focus are powerful solo. So far in testing it has seemed to me that the advantage of having a single HP pool massively outweighs any debuff advantage in most cases.
I finally had to give up on making a system that would work across all party sizes and encounter types. My final attempt works as follows: Each player has a Encounter Strength (ES) of Level + 6 and those are added together to determine the ES of the party. I then model a monster fighting two players and figure out at which level the fight is even just based on damage dealt. Then I calculate the monster ES based on the ES of two players at the given level. This is all very reasonable but then comes in the kludge - any monster encountered alone should be considered to have a ES 25% higher than listed because of the previously mentioned single HP pool factor. I mashed a bunch of encounters through my simulator and it seems to work out all right but it continues to irk me that I have not found a simple, clean solution. I suspect that unfortunately such a solution does not exist as otherwise some edition of DnD would probably be using it.
I don't like having monsters be so artificial as they are in 4th ed. even though it is very convenient for the GM. I also don't want them to be as ridiculous as 3rd ed. or Pathfinder where 'appropriate' encounters ended on round 1 and it was easy to be unhittable or unmissable. The old days of monsters being incredibly random and oscillating between lethal and trivial was entertaining at times but eventually unsatisfying and the new model defies immersion. If the perfect system does not exist though I will lean towards old school because I am convinced I can do a better job of the numbers than Gygax did back in the day.