One could make that case that on a nice spring day, younger voters are less likely to be at home answering the phone and presumably, an actual exit poll has a much lower margin for error since it gets a better (and larger) sample of actual voters, but simple common sense suggests an error in the IPSOS age cross-tabs.
