Joe has outlined everything up above BTW, at least to what I can think of.
botany all over again!
The plesiomorphic characters of subfamily Cypripedioideae include elongate stems, trilocular, and many plicate pubescent leaves. This makes Selenipedium the most basal of the group.
From the Selenipedium evolve the ancestors of Cypripediums. Cypriediums generally have shorter leaf internodes than the Seledipedium. Cypripediums are unilocular, another step in evolution. The most derived Cypripediums have reduced stems and few leaves, such as Cyp. acaule. (I bet something like cypripedium lentiginosum are the most derived of the cyps.)
As the Slipper species become more derived they inherit conduplicate foliage and reduced stems. These synapomorphies are seen in Paph and Phrags. These two genera are closely related, which is why the traits are similar. Their ancestor likely separated and evolved as they became geographically isolated.
Phrags probably have a reversal in locule characteristics and therefore are trilocular. The paphs remain unilocular, a trait from the Cypripediums.
And mexipedium evolved from the ancestors of today's Phragmipedium somewhere along the line (Either it retained the unilocule from the Cypripediums, or its another reversal into a unilocular carpel in progress? We don't know the ancestor of the Mexipedium and Phragmipediums, who are sister to one another).
Aren't Paph chromosome sizes much larger than Phrag as well.
...I hate it when someone labels a photo of some hybrid of paph philippinense as Phrag caudatum. I have seen that a few times!