Four big factors in your results...
About 10 years ago, I was using HDV, which is not only 4:2:0, but is actually 1440 pixels wide, stretched to 1920 wide. At the time, I wasn't always happy with the results, though recently I went back and revisited it with more experience and better tools and I was actually happy with the results.
Four things to consider:
Light things well. If your green screen is unevenly lit or has wrinkles or different colors in it, you're starting with problems that 4:2:0 will make much harder to fix. Work harder to get it right and that work will provide benefits throughout the rest of the process. (This also includes the foreground -- though watch spill and shadows on the background. If the foreground is too bright, you'll end up compensating and making the background too dark, and vice-versa, and too bright causes saturation problems and too dark creates noise and saturation problems.)
Avoid spill and reflection on the subject. Make sure the subject is far enough from the green screen and perhaps has some rim lighting to cancel any green spill. Also, try to avoid shiny materials near edges. For example, I recently did something that involved a puppet that had whiskers that were really reflective and reflected a lot of the green. Of course, this is difficult if you're on a budget and have a smaller screen and fewer lights. Also make sure you don't have green (if using green screen) in the foreground. Watch clothing, etc, that is in the same area of the color wheel as your green screen.
If you have flexibility of background, take advantage of that to help hide border issues. You're likely to get black, white, gray borders after keying, so depending on what's happening, if the background can help hide this, that's helpful. (Of course, if you have a specific background called out in advance -- a particular photo of a particular location rather than a kind of location -- you can't do this.)
Use the best software for keying that you can, and don't expect a single key to handle it all. You may use one key for someone's head, another for their body, and various garbage and solid masks to control things. It's tempting to keep tweaking keying controls or switching keyers (if you have options) to try to get the Perfect Key (tm), but in reality you will need to use multiple layers and tools.
Again, if your input quality (screen lighting, etc) is lower even a dozen layers and effects will probably not salvage it.
Bonus point 5: Test everything. Set things up, can you truly get even lighting on your green screen that's sufficiently bright and even? Check it on a scope, or pull some footage into your program and look at scopes. Document that setup exactly so you can reproduce it. Check clothing and props on the taping day to look for that lime-colored belt that the actor wore thinking "It's not green" or not thinking at all.
You'll probably get reasonable results.
[EDIT: I jumped into an answer before reading everything. 4K footage -- if your delivery is at 720p or 1080p -- covers a lot of problems. As stib says, downsampling essentially gets you 4:2:2.]