The downside of the unusual name is when the original record was, say, a US census record (so created based on oral transmission of the name): it may have been written not-the-way-you-expect and the odds of it being both written and transcribed the way you expect are only worse.
And so, my Abbenhouses: Offenhouse, Ebenhans, Allenbouse. You think I'm making this up? I really wish I was. Offenhouse was C Phil's WWI draft card -- his brother's was spelled and transcribed right and _they were living in the same town_. Ebenhans was Henry's family in an earlier census. Allenbouse was C Phil's family in a later census.
I've made enormous progress in tracking these guys down, but I'm unable to find either family in the 1920 census. I know they were all in Newport, Washington a few years earlier (and I've read all the pages of Newport, WA and Oldtown, ID formerly known as Newport, ID, in hopes they were there -- alas, they were not). By 1930, C Phil's family is in Seattle and the Ebenhans's are in "White River, King County", which we think is north of Kent. I have reason to believe that Henry's family was in Skagit in the 1920s: that's where they were before and after the Idaho years and that's where my grandmother would meet and marry my grandfather (which is why she isn't with her father in the 1930 census), after she graduated from Mt Vernon High School in 1926.
There's a small but real chance that both families were un-enumerated. I hate to give up, tho.