Is theory really that necessary?
What about playing chess? Is tons of theory of variations that you'll see twice in your life that overwhelmingly desirable?
In my youth, I wasted many hours learning openings while I should have been studying tactics and endings. Ten years ago I shifted focus. My OTB rating shot up from middling C Class to strong A Class. Now, I need serious work on openings to make Expert, and even to maintain my status as strong A Class.
However, endings and tactics remain critical too. I still win games after blowing an opening. I still lose games after getting an advantage in the opening.
History says you will never become any better than Capablanca unless you spend lots of time learning theory.
The kinds of miniscule advantages that professional chess players get out of an opening are almost irrelevant to amateurs. Some += evaluations are just because White has the two bishops or slightly more space, amateur players think += means White is winning or has the potential for a crushing attack. Unless you want to become a titled player just play moves and then check them against 'theory' later, even then you might not know why your move is considered worse than the main line!
Do you open with 1. d4 or 1. e4 usually? Do you have an idea of what move you like to play for your opponent's most common responses? Maybe you really like 1. d4 Nf6 2. c4 because that's a nice flexible move, takes control of some important squares, and you know hte positions that arises out of that series of moves pretty well.
Guess what -- you're doing theory.
It can not be avoided.
As to "which is better," you have to ask "better for what?"
For someone under 2000 if you want to improve (and you mean "helps achieve the goal" as a synonym for 'best') then the best thing you can probably do is avoid too much blitz, play serious long games with an eye to avoiding simple mistakes, analyze your own games (including openings) as deeply as you can, get a stronger player to look over your analysis and help you correct it when you go wrong, study tactics and endgames, and so forth.
If you mean "have the most fun," then it depends on what you find fun. Some folks really enjoy learning opening theory, and there's nothing wrong with that.
agreed. i blew 200 blitz points in a day or so. I am a 1550 USCF, and am at 1050 here on blitz after being 13xx. It happens. If youre anything like me, you will need theory for END GAMES more than anything. Either way you need not to fall for traps, and that is all theory.
Look at it this way: you could play thousands of games, learning by doing; and if you manage to remember, learn from and correct all your mistakes, playing every common position better the 2nd or 3rd time it turns up, your openings will start to approach what is called opening theory.
That´s one time-consuming method, and it needs a perfect memory and very good analytical capablities to make it work. The other method is just to start learning a few lines and variations, which you can do in say 30 minutes a day at the start. Take your pick! 
You don't need to learn a lot, just assure yourself you have a playable middle game out of the opening. If you get absolutely stuck one game you will notice this yourself. Then you should look up the theory to look for improvements.
I was attacking opening theory, really. I am much more supportive of middle and endgame theory. I have also played a lot outside of this site. Also, as someone else said, the logic behind moves is most important, especially for amateurs. For example, my favorite opening was invented before I started using it. However, I discovered it on my own. Also, I have heard that GM Pal Banko didn't study much opening theory, but endgame theory, which I am much more supportive of.
However, as you get closer and closer to that mystical 2000 level, even within several hundred points of it, knowing only that you can choose Qc2 or f3 isn't good enough. You need to know what you are trying to achieve with that move, and you need to know how to follow up with that choice.
The notion that opening principals are seperable from opening theory is silly. This example is a great example. General opening principals tell us to not move the queen early in the opening, but here we are moving it on move 4 . . .
Kingpatzer wrote:
However, as you get closer and closer to that mystical 2000 level, even within several hundred points
of it, knowing only that you can choose Qc2 or f3 isn't good enough. You need to know what you are trying to achieve with that move, and you need to know how to follow up with that choice. The notion that opening
principals are seperable from opening theory is silly. This example is a great example. General opening
principals tell us to not move the queen early in the opening, but here we are moving it on move 4 . . .
You misspelled separable. Also, while they are not totally separable, prInciples is the driving force between the actual theory. The principles create the theory. If you know the principles, your openings will be alright most of the time.