This is a really important question, and it's one that isn't asked enough.
From reading your question, though, it sounds like you are rushing a bit. Renowned physicist and biologist Uri Alon (34k+ citations) has written an excellent paper on exactly this topic'How To Choose a Good Scientific Problem' and one of his main points is that one, however cliché it sounds, should take ones time to conceive a good project.
I strongly suggest you read his paper. It's not exhaustingly long, and reads like a good talk. He also has a TED talk that challenges the 'A to B' dogma in science, which has helped me a lot.