|Image is from a chicken rearing forum, and is copyright its poster, user stephanie1992|
I did have it asked of me whether the splotch on her face could be caused by the masking gene, but I find this highly doubtful. If this dog was a piebald, the mask would be affected by the white spotting gene just as any other gene would be.
There are several possibilities as to what could be causing this dog's interesting coloring. It could just be an exceptionally unusual variation of piebalding, which I doubt. The facial markings are far too different from what would be expected in a piebald dog. It could be merle combined with the Harlequin gene, which I also doubt because the Harlequin gene appears to be isolated to great Danes. There is vitiligo, which causes the loss of color pigment from the body, but this dog does not look like a typical animal with vitiligo. This leaves only one option: the phenotype is caused by some form of mosaicism. This is a phenomenon that is seen in dogs on occasion, with the most commonly seen version being black spots appearing in otherwise red dogs (see here, here, here, and here).
Mosaicism, by definition, is when body cells have or express different genetic material. This can be caused by several phenomena. There is somatic mutation, which is where some mutation occurs in a cell after an embryo begins to develop, producing an individual that has patches of cells with minor differences. I suspect that this is the cause of such things as what we call birthmarks. There is also Barr body formation, where an individual must have two or more X's in their sex chromosomes. The body only wants one X chromosome activated, and as such will deactivate extra X's in the body's cells, creating a Barr body. This causes the tortoiseshell phenotype in cats as the basic red and black genes in that species are X-linked; it is unlikely to affect the coloring in other animals. Lastly, there is chimerism. This is where two embryos become fused into one early in development, producing an individual with two distinctly different genotypes that would be as similar as any siblings would be.
For this dog, she is likely either a chimera expressing a dominant black and some form of white (most likely extreme white piebald or extremely pale recessive red) or a black and white piebald with a mosaic spot on her face. I believe the first is the most likely, as the white in her coat is tinged with cream in a similar pattern to what would be expected in a very pale red, and I think the black is too extensive for a simple somatic mutation. This would mean that the two genotypes present would be dominant (K-) or recessive (aa) black and the pale recessive red (ee). It is also possible that some form of white spotting is in effect.
As for her breed, I suspect that she is a German shepherd crossed with a border collie, possibly more than half German shepherd as she seems very shepherd-y to me (but then again, she could just be a working border). Whatever she is and whatever her genotype, she is a very interesting-looking dog.