I'm leaning towards CF-WF too because of how light it is. Which is why I was having a hard time figuring it out. From what I was reading of what srtiels wrote, CF and PF are co-dominant, but both show visual when paired with WF (basically she says that CF acts the same way PF does when paired with WF.) Poor WF is just the odd man out.It's possible that this bird is not homozygous creamface, meaning that he has two copies of the creamface gene. His face is so pale that I'm going to guess that he has at least one creamface gene, but it's possible that he's CF-WF rather than CF-CF. He might even be PF-CF or PF-WF. If you breed him with a WF hen and they produce any pure WF babies, that will be proof that he has one copy of the WF gene and it's paired with either CF or PF.
If I understand correctly, you get a blended effect when these genes are mixed together. WF causes zero yellow/orange and the others are more than zero, so you're always going to get some color and it's going to look more like the "more than zero" mutation than it does like the "zero color" mutation.Poor WF is just the odd man out.