In the study of consciousness, neuroscience observes mental phenomena through physical correlations, using techniques such as fMRI, PET and EEG. These are considered valid and reproductible, and can be used to validate theories about the brain. Easterns also study consciousness but through a phenomenological method, the meditation, observing mental phenomena directly in first person, and producing many theories as well. Modern science critics point out that meditation also is a valid instrument to falsify theories about consciousness such as a microscope is valid to falsify a theory about a cell. Is it a valid claim?
Various branches of cognitive science (psychophysics especially, but not only) have demonstrated that we can be profoundly misled in our subjective interpretation of an experience. One of the most profound: when you look around, it seems as though you're seeing things the whole time. You're not. The visual stream is effectively blanked during a saccade where your eyes jump from one local target to another.
Given this scale of error, we can't conclude very much from reports of meditation except possibly that people tend to give consistent reports (or not, if they don't). You can't get off the ground scientifically with meditation alone because it is in a regime where almost every introspection is likely to be misleading. It's not even reliable enough to falsify scientific theories unless there's a strong prediction of what people ought to report (which basically never happens--we don't know enough to even attempt such a thing).
However, it's perfectly okay to use meditation to alter your state of consciousness and then study that (using methods which are less subject to prediction, retrodiction, observer bias, and all the other things that our brains do automatically for us to provide an interpretable (if not always accurate) experience). Asking someone to meditate is not really any different then asking them to imagine a picture of a dog or to do any other internal mental task. Knowing how our subjective assessment of our state of consciousness ("I feel hyper-aware") maps onto activity patterns ("huh, anterior cingulate, and...") is a big part of what cognitive science should deliver to us, at least as a first rough draft of how consciousness works.
Well, as with almost everything in philosophy, it depends on the definition you use, in this case of "science".
Meditation is open to falsification. If you do it and experience something that leads you to a hypothesis, I may do it and experience something that falsify your hypothesis.
Notice that the fact that a phenomenological method is ontologically subjective is not the same as not being epistemologically objective. It is subjective because it is an interior experience. But it may be objective in the sense that when I do it and you do it we both share some features of the same experience.
Of course, this would dismiss most mythological accounts of the experiences, and should leave only the things that are common and consistent.
The problem here is when you take the empiricist part of science. Empiricism refers to experience, but in general there is a supposition of experience as only what comes through the senses. But if you take some form of radical empiricism stance, you may consider anything you experience as, well, experience.
The major example in this would be mathematics. It is common to refer to math as not science, but a subarea of logic and philosophy. I don't want to go into that discussion specifically. All I wanna do is point out that math has the same issues.
When you prove a math theorem, for example, what you do is an interior process that leads to an interior experience. The experience is the "aha, this proof makes sense!". Then, you go to your peers and say: check this proof. By this you are telling them the equivalent of: do the interior math process and check the interior experience to try and falsify my results.
So, again, it all depends on what stance you take on "science".
This is not an easy nor settled question. Personally, I think it doesn't make much sense to keep those kind of knowledge - another problematic word - so far apart.
I think the methodology of consciousness studying is the same of cognitive science methodology. The difficulty in considering for example the theories of meditation in Hinduism or Buddhism as scientific theories of consciousness is the same difficulty that is in the origin of the scientific methodology of the double-blind trial: How to eliminate subjective, unrecognized biases carried by an experiment's subjects and the conductors of scientific experiments? If the experiment's subjects and the conductors of scientific experiments are the same person the problem is compound. Justification of a scientific theory consists in appealing to something independent: Meditation's experiences ground the beliefs implied by the meditator's sincere reports of such experiences, provided they can be said to cause those beliefs. But it may well be that the beliefs are part of the cause of the experience rather than the other way round.