whether string theory meets the ... definition of "science."
This is a very good question. The short answer is yes. The
details are as follows:
1) It is hard to judge a theory "as a whole". A theory can be
a big thing. Even with the best of theories, some of what people
say on behalf of the theory might be wrong. For example, even
Newton's laws are often misstated, but that doesn't mean mechanics
is wrong or unscientific in any significant sense. So IMHO when
judging a theory as a whole, we should pay attention to the
"center of mass" of the theory, and not pay too much attention
to the fringes.
Conversely, keep in mind that even the wickedest of lies contain
many elements of truth.
2a) String theory is heavily constrained. By way of illustration,
consider the significance of 11-dimensional spacetime. As of
now, 11 a viable hypothesis, while 9, 111, and almost all other
numbers have been ruled out. Let's be clear: D=9 string
theory is not only falsifiable, it has been falsified. The
D=11 theory could have been falsified in the same way, but
it wasn't. This is exactly the criterion for being scientific:
the theory must expose itself to being tested, and it must
pass the test.
2b) The analysis of ID starts out the same, but ends up in a
completely different place. Some versions of ID are like D=9
string theory: they are not just falsifiable, but false. The
problem is that the other versions of ID do not pass the test;
they skive off the test entirely. The only way they can avoid
being falsified is to make themselves unfalsifiable, by
twisting into meaninglessness the definition of the key words
("intelligent" and "design").
===========================
The knock against string theory is that it doesn't tell us anything
we didn't already know, except in situations that cannot presently
be tested.
Even if it never told us anything new, that wouldn't make it wrong.
One might argue that the law of conservation of momentum doesn't
tell us much beyond what Newton's third law of motion already told
us, but that doesn't make conservation of momentum wrong or useless.
===============================
As a side-issue, we can consider not only the methods but the goals
and methods. The string theory guys are avowedly working toward
a consensus. In particular, ten years ago there was a D=11
version of string theory, and a seemingly-incompatible D=26
version. The discrepancy was resolved by M-theory, and everybody
was delighted.
In contrast, the stated goal of the ID guys is to stir up controversy.
They benefit, politically and financially, from the controversy. They
say we should "teach the controversy" ... but there should never have
been a controversy. It's like the controversy over whether the moon is
made of green cheese ... or flag burning ... or whether Carl Foreman
was a commie traitor. If you give me a few million dollars, sure, I
can create the appearance of a controversy on those issues, too.
ID is so obviously false (unless it is twisted to make it unfalsifiable)
that it is a scandal that we should even need to discuss it. But alas
we do need to discuss it. We do not, however, need to treat it on the
same footing as scientific theories.