The VCARD shape for phone and email is non-intuitive

See gitter chat at https://gitter.im/solid/app-development?at=5bc0696e3844923661276f51

Alfredo Torre @sentenza Oct 12 05:29
My email address shows up as a separate "node" and I dunno why this is happening:

:id1539184351555 a n:Home; n:value <mailto:myemailaddress@email.test>.

Could anyone explain what does this mean, please?

[...]

Alfredo Torre @sentenza Oct 12 05:31
but here I should see something like n:hasEmail, shouldn't I?

@melvincarvalho notes:

[..]

 :Home a owl:Class ;
    rdfs:label "Home"@en ;
    rdfs:comment "This implies that the property is related to an individual's personal life"@en ;
    rdfs:isDefinedBy <http://www.w3.org/2006/vcard/ns> ;
    rdfs:subClassOf :Type .

So in general the vcard pattern is

:me  vcard:hasEmail :foo .
    :foo  a vcard:Home; vcard:value <mailto:myemailaddress@email.test>.

or with bnodes

:me  vcard:hasEmail :[ a vcard:Home; vcard:value <mailto:myemailaddress@email.test>].

It is working as designed by the RDF VCARD group. The email you have is an intermediate object which has a type which is say Home or Work, and value which is the actual email.
This is not the way I modeled in in my own early contacts vocab — there you had a location, which is a Home or an Office or a Vacation and that location would have directly the phone number, address, etc.
I was chatting yesterday to @danbri and Vicky of Schema.org, and I think they said that in fact in Schema they did not put in the extra node, but had an approach more like the contacts

@prefix contact: http://www.w3.org/2000/10/swap/pim/contact# .

See https://www.w3.org/2000/10/swap/pim/contact.n3

A possibility would be to move solid to Schema.org in general and in this case it might make for a more intuitive structure. See for example

https://schema.org/ContactPoint

No source files, tests, or entry points are named. Start by reviewing the VCARD examples and the linked Schema.org and contacts vocabulary references, then determine whether the project should change its data model. Done would require an agreed design and an identified implementation scope.