We've talked about extracting strawberry DNA for use as carrier DNA in our Vegan Cheese project, just because it's such an easy and abundant DNA source. Haven't gotten around to experimenting with it though.
As far as I know, there is nothing special about salmon sperm DNA. You just want a heterologous DNA source that does not crossreact with and is easily distinguishable from the other DNA you're working with. That typically rules out human, mammalian, E. coli, yeast DNA, etc. Fish DNA fits the bill, but so does plant DNA.
Sperm cells are very small, so you actually get a lot of DNA per volume of cells. And fish sperm (often herring sperm in Europe) is a cheap and abundant source anywhere there's a fishery industry.
The use of salmon sperm DNA in molecular biology actually dates back all the way to the discovery of DNA by Friedrich Miescher. He was originally working with pus from bandages (neutrophils - another good source of DNA), but eventually switched to using salmon sperm from the large-scale salmon fisheries along the Rhine.
So the answers to "why salmon sperm DNA?" are: (1) at least it's not pus DNA, and (2) that's what we've used since the 1870's - if it ain't broke, don't fix it...
Patrik