Can (or How) Anthropic Principle be made into a scientific theory?

The statement you quote is a truism, so not much science in it. There are however two ways you can extract science from the anthropic principle.

One is that it can give you non-trivial constraints on parameters in current theories. In most cases the existence of intelligent life itself is too much of a detail that enters the argument. It's more that you'd be saying we have observed that the universe is such that e.g. carbon exists in such and such an abundance, and this is only possible if parameter x is in some specific range, so you've constrained your parameter. That constraint might be more or less interesting, but nothing unscientific about this.

Second one is that you might try to come up with a theory in which the "evolution of life" is well-defined in scientific terms, such that you can argue there's some optimization principle going on which results in some specific theory, that being the one we observe. However, there's some arguments that our standard model might not be the only one supporting the evolution of life (Google for: Weakless Universe). In any case, taking this road is going to involve a lot more than only physics because first you have to find a way to define "life" in mathematical terms.

The anthropic principle (at least in it's weak form) is neither philosophy nor science. It is simply a statement about the admissibility of evidence.

Any precondition to an observation is not evidence about the whatever you are observing.

Duh!