I've been deliberately putting this topic off to last because I'm not sure what to say about it.
It's not that solo monsters being more subject to the laws of Critical Existence Failure (http://tvtropes.org/pmwiki/pmwiki.php/Main/CriticalExistenceFailure) is somehow wrong-it's just that it's different, and that difference means that some of the common conceptions of how combat works are out of touch.
I have to confess, much of my theory on the topic of solo monsters is based on an arbitrary number someone told me that encounters last for five rounds. And that's about it. It's a completely arbitrary statistic that I don't have the numbers to back up, and yet it's going to found the basis of this essay.
Now, the reality is that no two parties are identical (unless, well, they're identical) and how any given group deals with combat encounters varies. A group with three wizards all tossing Icy Rays everywhere may have monsters drop three at a time but dealing little overall damage. But a group with two strikers is likely to take out one monster per round for the first few rounds while they burn through encounter powers. So, even though there's five monsters in a standard encounter, the longevity and effectiveness of those five monsters isn't a constant.
If we just assume that a solo monster has five standard monster level attacks, that's great-ignoring other action economy problems-but we run into the nasty problem that it's not indicative of how a standard encounter functions*; it's like if everyone was attacking a different creature and not focusing fire. This means that, in a five round combat, you'd expect to see the equivalence of twenty five creature rounds.
Compare this to an ordinary encounter, where there's five creatures in the first round, four in the second, three in the third, two in the forth, and then one straggler in the fifth. That's only fifteen creature rounds by comparison, so your solo monster is 66% more effective over the lifetime of an encounter!
On the other hand, perhaps not. Controllers can't use multi-target damage, meaning that the net damage per round being dealt by the party is reduced. This can partially be balanced by the fact that there is no chance for lost damage in the form of overkill. On the other hand, if you're playing with an Essentials Assassin, there's also no chance for the Death Attack feature to trigger! And finally, despite being worth five creatures, ever since Monster Manual 2, solo monsters are only considered to have 4x HP in order to avoid having fights drag on too long!
The fact of the matter is that, I don't have the hard data to come to a decisive conclusion here. If we assume striker overkill balances out lost controller damage, then the fact that a solo monster only has 80% HP of a standard encounter leads us to the conclusion that we can multiple those 25 creature rounds by 80%. That's still 20 creature rounds, or 33% effectiveness over a standard encounter. So, what's the solution?
If we assume that because it only has 80% HP the solo monster only requires four rounds to defeat, we can expect that the first round it will be unbloodied, and sometime around the middle or end of the second round it will be bloodied. Working from this, we can use a 3 Creature Round x2, plus 4 creature round x2 model; in short, if the creature's effectiveness increases by about 33% while bloodied, but its effectiveness pre-bloodied is only three creatures rather than four, we're somewhere around the proper level we're aiming for. But again, this is very ad hoc, and making a lot of unsupported assumptions. For instance, most multi-target attack type monsters can't effectively attack the entire party, but a creature with sufficiently large close burst powers can.
Also, again, I think a lot of this has to go back to the Game Theory of Solo Monster design. If the players know that solo monsters are more powerful bloodied than not bloodied, they will want a greater percentage of the fight to occur with the monster not being bloodied. Upon first instinct, you may think that doesn't make sense-after all, in addition to our anticipated values above being "Bloodied by the end of the second round," half of the creature's HP is half of it's HP, right?
Not really-the fact of the matter is that party damage is wildly different between at-will output, encounter nova output, and daily nova output. If players know that the creature is more deadly while bloodied, then players can reserve powers for after the creature is bloodied. Of course, if the solo monster is aware that because of this PCs are taking it easy, the solo monster can use that information as well, so it gets complicated. But in reality, it's been my experience that most PCs don't think this way-they like to burn through encounters.
And to a large extent, that's wise. Not every encounter ends with every encounter resource having been spent, and for creatures with regeneration, every round you procrastinate is a round of additional HP you have to deal with.
Other alternatives for dealing with this kind of playstyle are having a creature deal low damage but have extremely high defenses-typically in conjunction with regeneration-for the first half of the battle. Each round is relatively non-damaging to the party, but still a cost in resources; procrastination only works in the monster's favor. Once the monster becomes bloodied, it becomes truly effective, but the party may have to expend its novas in order to get it that far, or risk being bogged down. The major downside of this gameplay style is that it's time consuming, and has a risk of alienating players, regardless of whether it makes for dynamic combat.