The cell phone effect doesn't happen mainly because the difference in voting between landline youngs and cell phone youngs is minimal. It's not as if only conservative younguns have landlines and only liberals have cell phones. It just depends from person to person and I doubt it is political.
But it results in the young vote being undersampled.
I don't really see the young vote being undersampled. Most voters make sure they have an appropriate amount of youngs in their sample. The real question is whether landline youngs vote differently than cell phone youngs. I say the answer is no.