Another definition:
out-cross
[v. out-kraws, -kros; n. out-kraws, -kros]
verb (used with object)
1. to cross (animals or plants) by breeding individuals of different strains but, usually, of the same breed.
2.to produce (a hybrid) by this method.
noun
3.a hybrid animal or plant so produced.
4. an act of outcrossing.
I am pretty sure any recessive trait will be propagated by line breeding. The chance of two alleles being the same in two related animals is much higher. In fact all of those morphs were more then likely line bred to get a certain look. This is the reason you don't find many of these traits in the wild. In nature populations are constantly outcrossed and unless a recessive gene is useful and lucky it will be more than likely lost.
In a genetics sense, the term is
properly defined as Jereme stated. Hybridization can come into play, but again the term is more valued towards introducing genes within the same species. The
idea and concept is that by outcrossing to an unrelated individual that you are increasing genetic diversity all the while
assumingly reducing disease and/or "negative" genetic anomalies. The practice is used in line breeding in
attempting to restore breeding vigor. It should be noted that line breeding is
different from inbreeding. Line breeding results in a line sharing a common ancestor. Inbreeding is, essentially, a more dramatic form of line breeding where the common ancestor are parents and the offspring are more closely related.
A misconception is that inbreeding only brings about recessive traits and that recessive traits are "negative" or undesired traits. Dominant traits can also be perpetuated through inbreeding and can also be of negative or undesirable results, e.g., achondroplasia in humans, an autosomal dominant dwarfism mutation. In this case, being homozygous recessive results in "normal" height and limb-length, and even 1 copy of the mutation, i.e., being "het" is enough to suffer from the disorder.
In contrast to your last statement, many corn mutation are actually wild-derived. The first albino corn was road cruised circa 1953. There are natural populations of anerythristics (all 3 types) in FL, with anery Type A originating from around Fort Meyers if I recall correctly. The first caramel corn, again if I recall correctly, was a wild caught animal in a FL pet store that Rich Zulchowski stopped in and purchased. The palmetto corn was a wild caught. Same goes for a number of other "non-natural" mutations of other species (albino harquahala rosy boas, albino whitewater rosy boa, albino California kingsnakes, etc.) that were wild occurring.
Food for thought... Outcrossing can also result in outbreeding depression, where the outcrossed offspring actually result in less vigor/fitness than those offspring from line breeding or even inbreeding. So just because you outcross doesn't always mean you're strengthening lines.
You'd be surprised how related wild populations can be, particularly if they have a small range to begin with or of an isolated population. I believe there was an eastern box turtle study recently documenting this.
Additional food for thought, this very subject is why I rarely breed homozygous recessive to homozygous recessive (same for homo. dominant). I typically breed a heterozygote to a homozygote (example, my current male breeder tessera is het for butter/amber stripe and one of the females he gets paired to is a full homozygous recessive butter stripe). I may produce lots of hets, but heterozygosity can also help in keeping a line strong.