I thought not, and not merely because of the difficulty of carrying it out.
We all know that that kind of thing is what actually happens in the real
world. The fact is that we can, in large measure, specify what "fitness"
consists of for any ecological niche, and we can point out that not only
will polar bears not do well, *as* polar bears, in the Sahara Desert, but
that camels won't do well if they suddenly find themselves in polar bear
lands. Camels are not very good at catching fish, etc., so they will not
last long.
Evolutionary theory claims that variations occur. Is this tautological?
No.
Evolutionary theory claims that some variations are less suited to survival
and reproduction than others. Is this tautological?
No. This is obvious. If a polar bear gives birth to a camel, that camel's
chances of surviving without human help are pretty slim.
So, just where is the tautology supposed to be?
It seems to be in the weird idea that fitness is *defined* simply in terms
of survival. But it isn't. Fitness is defined in terms of what an organism's
traits are and what the predictable consequences are of having those
characteristics. This is not always *easy* to test, because trade-offs can
be subtle, and we don't usually have very good information upon which to
make predictions. But, to argue that it is untestable at all, or, worse,
that it is untestable even in principle, is simply ignorance run amuck
(i.e., Phillip Johnson, et al).