EzekielRaiden
Follower of the Way
Only if they mix well, meaning two classes that share a prime stat (here, Int is the core stat for Wizard and the subclass stat for certain Warlords) and also, at least in the vast majority of cases, do different but complementary roles (as is the case here, where you have Controller/Leader--"set 'em up, knock 'em down.")
Another example of this in action is Paladin|Warlock, since a single feat can then pick you up Plate proficiency, making you an extremely beefy, support-heavy Striker. Shaman was IIRC also a nice option in general because feats could let you use your spirit companion without some of the annoying limitations imposed by hybrid rules. But try hybridizing something like Barbarian|Wizard or Paladin|Swordmage and you'll often find yourself weaker than the sum of your parts.
Multiclassing was definitely a thorny problem for 4e. They wanted to fix the common complaint that MC was both too powerful and too weak, that it was often a trap unless it was brokenly good, and that it usually required massive preparation to turn to useful ends. They ultimately met that goal, but the thing they produced was still vulnerable to the "crappy choices" problem (moreso for hybrids), and the end result was pretty steep cost for not a lot of gain (especially paragon MC.)
Were I in charge of a 4e retrofit retroclone, something meant to fix genuine flaws in the ruleset, I'd definitely be putting a solid chunk of my budget toward reworking muticlassing. It needs to have a reasonable cost, but remain flexible; it needs to be clear about how to use it well, while still functionally supporting the off-the-wall choices that might not be the strongest option.
Perhaps, with the other tweaks I would be making, I might have "paragon' multiclassing instead be something that eats up your Heroic Origin. That's my name for my idea of merging Backgrounds and Themes into a single, cohesive choice made at 1st level alongside Class and Race. Do something like:
"When you take an MC feat, you can choose to replace your Heroic Origin with Determined Dabbler, which gives a taste of another class's powers and features. Once you reach 11th level, you have access to all the basic features of your second class. If you also acquire a MC Feat for that class, you can swap one of your powers for a power from that class. By taking a Paragon Training feat for that class [read: "advanced" MC feat, must be level 11 and have an MC feat for that class], you can swap one at-will, one encounter, and one utility power for choices offered by your second class, and you use your highest ability modifier for hit bonus on these attacks. If their effects call for a second ability score, use your second-highest score.”
You won't pick up "build" (what 5e calls "subclass") features, so you'll never be quite as good as a proper Wizard or whatever, but you'll have the heart of it. And with MC feats being quite strong, then as long as Paragon Training feats are also a solid option (perhaps granting a second skill training and a class feature?), this would make the "paragon MC" option reasonably costly (eating up the real 4e equivalent of your Background, Theme, and two feats) while still supporting whatever things most players would want to do.
Obviously just spitballing. But I think there's at least the spirit of a useful rework for 4e-style MC there. And then Hybrid would fill in any remaining "full-on mix two classes together" hankerings. Between that and the reasonably comprehensive list of classes (my only desire being the addition of a PF-style Summoner), you'd have a solid structure for supporting variety while conclusively eliminating the "plan everything to the third decimal place" problem.
