If you know it's fuzzy, then you should be able to tell me what's fuzzy about it. I say there's nothing fuzzy about it. I can't point out to you what's not fuzzy because I see nothing fuzzy to point to.
You're making an assertion, but some substance behind it.
It's fuzzy precisely because it's not well-defined. If you were to give me a precise definition you would de facto dispel any notion of fuzziness - assuming your definition is precise enough and can stand up to the YECist idea that only God can produce new kinds.
To directly answer your question, I think these two quotes sum up the difficulties I see in your position:
The asserted boundaries between the kinds -- the position that the kinds are unrelated -- is arguably the most divergent view of creationists from mainstream biology. Those challenging creation biology often ask what basis creationists have for asserting that such boundaries exist, or for determining what those boundaries are.
The project of determining the precise boundaries between the kinds is not agreed upon by creationists. Creationists generally assert that conclusions about common ancestry should only be drawn if there is "substantial evidence" to support the conclusion. As to what qualifies as "substantial evidence", creationists are often at odds with each other. However, they are in unanimous agreement that humans and other extant primates are not in the same kind.
In the absence of the ability to directly observe life in its original form, classification of kinds generally revolves around reproductive compatibility -- that is, created kinds are generally seen as having common descent if they are reproductively compatible.
The classification is more difficult when reproductive compatibility is partial, as in the case of the
mule, a
hybrid of the
horse and the
donkey which, although viable, is not fertile. While it is possible that the two species descend from a common ancestor due to their reproductive compatibility, it is also possible that they do not, but were created separately with reproductive systems similar enough to create viable offspring, but not similar enough to create fertile offspring.
http://en.wikipedia.org/wiki/Created_kind#Boundaries_between_kinds
and:
According to the modern theory of genetics (which the creationists say they accept), evolution takes place through the natural selection of variations brought about by genetic mutations. By postulating that there are certain limits beyond which mutations cannot proceed, the creationists are in essence claiming that there is some mechanism, whether biochemical or biomechanical, which only allows certain mutations to appear (those within the limits of the "created kind"), and rigorously excludes certain other mutations (those which would carry the organism outside these limits).
...
Hence, some creationists have now dropped the requirement of "interfertility", and have asserted that any organisms that can breed with each other and produce offspring, whether fertile or not, constitute a "kind":
"Creationists have long felt a need for a classification that would include in one consistent category all organisms that interbreed under any conditions." (David Menton, "Species, Speciation and the Genesis Kind", Missouri Association for Creation, October 1994)
This definition, however, also produces problems. In the northeastern United States, for example, are found two species of tree frogs,
Hyla versicolor and
Hyla chrysoscelis. The two are absolutely identical in appearence, and the only way to distinguish them in the field is by their slightly differing mating calls. One of these species is a "polyploid" of the other, that is, it developed from the other species when a chromosomal abnormality left some individuals with twice the normal number of chromosomes. (Polyploidy is a very common means of plants to produce new species--in fact, most domesticated food plants like wheat and rye are polyploids--but is comparitively rare among animals.) There is no doubt that the two frogs share an ancestor/descendent relationship, and that one evolved from the other through polyploidy.
For the creationists to consider these two virtually identical frogs as being of different "kinds" would be absurd on the face of it, since they are so alike they can be distinguished only in the lab, and they obviously share evolutionary descent. So naturally, the creationists would like to lump these two species together as "variations" within one "created kind". But there is a problem for the creationists--the two
Hyla species do not, and, because of their chromosomal differences, cannot, interbreed. Not only do they not produce any fertile offspring--they are incapable of producing any offspring at all. The same problem arises in connection with plants--the polyploid descendents of particular plants can no longer produce viable seeds with the parent stock, and thus cannot produce any offspring with the parent species. Therefore, the creationist, using the criterion of "interbreeding", must conclude that the two are different "kinds", even though one is obviously a descendent of the other (polyploid plants have been successfully produced and bred in the laboratory--in fact many of our food crops are polyploid descendents of corn and wheat plants which can no longer interbreed with the parent stock).
Once again, the creationists must either admit the existence of evolution between "kinds", or they must change their definition of what constitutes a "kind". Thus, we are finally led to:
"If two organisms breed, even though it is infrequent, they are of the same kind; if they don't breed but are clearly of the same morphological type, they are of the same kind, by the logic of the axiom which states two things equal to the same thing are equal to each other." (Wysong, cited in Kitcher, 1982, p. 152)
One may dispute just how "logical" Wysong's definition is (on the one hand, organisms which interbreed are of the same "kind"; on the other hand, organisms that
don't interbreed are also of the same "kind" if they look enough alike), but there is no disputing that even this loose definition causes problems for the creationists. Now we need to define what constitutes an organism "of the same morphological type". Gish points out, "The division into kinds is easier the more the divergence observed." (Gish, 1978, p. 35)
http://www.geocities.com/CapeCanaveral/Hangar/2437/kinds.htm
(the reason I employ a cut-and-paste is to save you time, and also to cut out any references to evolution you might straightaway reject. The material quoted deals exclusively with the concept of created kinds and does not need to assume the truth of evolution to read.)