Everyone and their brother-in-law around the writing-related blogoverse has been posting or linking to this cartoon, and yeah, it can be true, but sometimes it's not. It depends on the skill of the writer, really, as do so many things.
You can be an Anne McCaffrey and make up a lot of words and names which fit your world well and are internally consistent in their tone, derivation, etc., and incorporate them very smoothly into your text such that the reader is never confused about what's what. Or you can be a JR Ward whose made-up words and names sound like she hit up a local junior high for suggestions, only use your own made-up rules when you feel like it, and just toss the new vocabulary into your text however so your readers are always going, "Huh??" and having to flip to the glossary. Credit to her for putting a glossary in her books, and for putting it at the front of the book so you know it's there right when you start, but seriously, the Black Dagger Brotherhood books are an advanced course in how not to make up words and names, and how not to use them in a piece of fiction.
But it depends. It's one of those things which is very easy to mess up, so many writers do and many readers remember the trainwrecks. But the problem isn't with the concept of making up words; it's with the writers and how they do it. There are a lot of aspects of fiction writing -- including many of the subgenres and plot devices and techniques and whatnot which get massive snarking -- which really aren't problematic in and of themselves, but rather which are difficult to do well and so are rarely done well.
But it's like waxing sarcastic about Everest expeditions, just because 99.9% of the people on the planet don't have the skills or resources to actually make the summit. High-end mountain climbing isn't stupid; people who try to do it when they aren't prepared to do it successfully are stupid. Or maybe they're just still learning. But don't blame the mountain if most of the people who try to climb it end up failing. (At least with writing, failure rarely means death.)
I think deciding whether to try a new or difficult device or technique is one of those areas where a writer has to be brutally honest with him- or herself. It's easy to say, "Oh, I've seen this done, so I can do it too." Or "Well, I know a lot of people say you shouldn't do this, but Chris Awesomewriter did it and it was great so what do they all know?" It's harder, though, to make an honest assessment of whether your particular skills are up to the task. Am I honestly as good as Chris Awesomewriter? If not, maybe I should back off on using that one device Chris used to such good effect, but which a hundred other writers have crashed and burned on.
Which isn't to say one should never try new things. I try new techniques and plot devices and character types and narrative voices all the time; it's one of the reasons I have so many WIPs on my hard drive. I just don't share them all, because I've developed a decent sense of when something's working and when it's not.
There's something to be said for a practice piece, or what an artist would call a study. Labelling something as being For Practice means the pressure is off. You don't have to worry about whether it'll work or whether it'll be perfect or whether that train will wreck and take the station with it. If it's just an experiment, then you're free to fail and to learn from the failure and go on to the next piece.
I get uncomfortable, though, when something as basic as making up new vocabulary for an SF or fantasy story is mocked and held up as something which makes a story suck. Only bad writing can make a story suck.
The best way to become a writer who doesn't suck is to practice a lot, try new things, and learn from your failures. Maybe after three or four or a dozen failed expeditions you'll finally make the summit, while the folks who carefully avoid everything that's difficult and therefore prone to failure never make it past the foothills.