Add X25519 Subkey for Ed25519 Test Vector

[TJ-91]

Currently we only have a PQ/T encryption subkey for the A.1. test vector. Wouldn't it make sense to add an X25519 subkey?

[falko-strenzke]

For what reason do you want to add the traditional subkey?

[TJ-91]

I suppose the test vector is meant as an illustration of the case that is described in 8.2. Key generation strategies (Point 2).

In a typical scenario, the owner of the certificate would like to enable PQ(/T) encryption, while remaining compatible with PQ(/T)-incapable clients. Not having a traditional encryption subkey limits the general usability of the certificate for non-PQ(/T) clients. It is therefore not something that I would expect in a typical certificate.

While not critical, I think it's good if the test vectors reflect what we expect to see in the wild.

[wussler]

Also switch the same test vector to use PKESK v3 and SEIPD v1 for completeness