The fun thing is that the biggest problem with multiclassing is that it's too easy to accidentally make a underperforming character.
Define "underperforming". And while you are at it, if you could please explain its practical impact to play, if any.
Unlike 3.x or pathfinder, 5e multiclassing is not the "route to power".
Or, there is no spoon. This is the kind of claims made by folks who go around warning of "traps" hidden in 5e. You know what I mean, terrible choices you can make that ruin your character because they make you suck. Just another spoon, IYAM.
Playing AL at my FLGS when 5e was newish, it was funny how many people tried to make min/maxed multiclassed characters only to find themselves lagging behind generic single-classed characters.
Again, you seem to be using a very specific, and narrow, definition of "lagging behind". Because, to me at least, its an exchange. You are trading increased effectiveness with a narrow skillset for flexibility. You clearly look down on that transaction. That doesn't make it objectively weaker. Just seeking something different than what you value.
Don't miss ASIs. If you invest 3 levels, or 7 levels, invent the next and get your ASI/feat. It really hurts when you don't.
No. It doesn't. Not really. Look, I get what you are thinking is a thing. The obvious bump provided by the ASI/feat. But lets say a 3rd level BM fighter has two choices: Take a 4th level in fighter for that ASI/feat. Or he can take a level in rogue. You contend that it "really hurts" not to take the ASI. For what? Are you one of those folks who insist you need to raise your stat to 20 before considering anything else? Cuz that's been shown to be untrue, thanks to bounded accuracy. Are you saying the opportunity to take a feat is a "must"? Well, the way I see it, the fighter who takes that rogue level effectively took a "feat" that gave him expertise in two skills, 1d6 sneak attack, and a secret language. Sounds like a pretty decent feat comparable to any other on the feat list. Is it postponing his 5th level fighter level for the extra attack? Of course. And that *is* a thing, I give you that. But its a conscious decision to delay that future increase in combat effectiveness for flexibility. You discount that, I get it. But many don't. And evidently neither do the devs. And, experience playing 5e has shown me, neither does practical play at the table.
I for one love that 5e multiclassing is actually interesting and full of hard choices. That's what informs me that is is a good subsystem. Every time you consider doing so, you know you are giving up something to get something else. Without that exchange, it would be, as you point out, a "path of power" akin to those other editions/systems. Which it clearly isn't. And since MC characters continue to contribute to play just fine every day across thousands of tables, you contention that it is "weak" is lacking practical evidence.
Don't miss out on your 5th level (or 11th level) power bump for long. Most classes have a serious bump at 5th and a smaller one at 11th.
And what are you claiming will happen if they don't take your advice? Is the character leading the party to a TPK because they missed the boat on the power bump? Is that your contention? Because I've played in plenty of games where a PC's MC path postponed those "power bumps" for multiple levels. Game played fine.
Now, fighter is one of the more cherry-pick-able classes (cleric 1 being another one) so they wont' be too bad off early on.
More unprovable personal opinion disguised as objective fact. Wow, but is that untrue.
But Rogue 3/Fighter 2 is weaker then Rogue 5 or Fighter 5. In additon to either of those having +2 to an ability or haivng a feat...
I already covered this above. The things you get for one are comparable to, sometimes greater than, what you can get from a single ASI/feat.
...Rogue 5 has Uncanny Dodge gives half damage to one attack every single round and an extra d6 of SA, while Fighter 5 has a subclass and gets two attacks (three with two-weapon fighting). It's not badly off, still within the curve of classes.
I will defend that Rogue 3/Fighter 2 is "withing the curve of classes". As long as, what you are trying for consists of things you get for taking those levels. I get that your infatuation with narrow power curves, within tightly defined categories, are paramount to you. But I don't think you are writing a book on what constitutes playing 5e as intended*.
(*something not only self-evident by the rules as they are written to work as they do, but that they can be played that way and the game continues to achieve its stated goal.)