Programmers care about efficiency - it's just that we were born this way :) When you create text RPG - it of course depends on your definition of RPG, but I'll go with mine, which goes like "Game where you can go through various locations, meet NPCs or monsters, fight, talk, have equipment, develop your character, world is presented to player in text form, and communication is based on text commands" - you will notice it takes more then few "ifs" to get things done. You may create really really simple RPG, but you will see that without proper loop and structure of game, it's really hard to change anything - as making one change/addition affects the whole, most likely creating many bugs.
I think that the most important thing when you are learning to create games, is to write games, but not in purpose of making it really good game, but in purpose of learning. If you made buggy snake, but it works, and you learned something - it's okay. This snake can be counted as successfull project. It does not matter that sometimes it hangs, that you have no menu, that after loss game just quits - you learned something, and that's what counts.
If you decide to develop your project further, you will also learn -at least that's my guess; you will learn that game is more complicated to create then writing few "ifs", and that game structure is really important.
They say that when you want to write program, you should start with pen and paper. It's true - if you organize your program at the beginning , you may see the problems before they are too hard to fix. First game I created, Tic-Tac-Toe was terribly designed. But I learned something, and it worked, so it's a success. Next games were better organized, and I was learning something new every new project.
As to your quote:
I honestly thought making a game like tetris, snake, etc. was a lot harder than making a text-rpg that just uses some if-statements and a story.
I also did. But because it does not involve graphics, it does not mean it is easier. Snake can be made basing on tiles, so collision detection will require no maths but simple comparison, and creating classes for snake is much simpler then for RPG.
I am telling you what I think basing on my experience. You may do whatever you want. I abandoned my first text RPG, and I consider it a good decision. I might have get it working, but I would design things completely wrong, hence I would not learn too much - I would waste my time on looking for workarounds of problems that would never be there if I used correct patterns in the first place.
Anyway, good luck. I hope you succeed in whatever you do.