Just go and read RFC4026 [ietf.org].
You have linked to a non-standard document, which is not a normative reference.
No consensus required to publish. Non-standard use of the designation "Virtual Private Network".
Sample a large enough number of the population, and you are bound to find people abusing terms.
Their usage is inaccurate. Informational level RFCs don't have the blessings of the IESG or the community.
The fact that you use the term "secure" in this context already shows that you are clueless. What do you mean by secure? Security can include any of the following
Are you trying to be a smart-ass?
All three are necessary conditions.
A Virtual Private Network is the combination of the concept of a Virtual IP Network with a Private Network.
All those various solutions are virtual networks. Virtual networks are networks that have been extended over the top of another physical network --- in other words, the network is tunneled or encapsulated, so the IP addressing of the network is not that of the physical network.
The important bit is Private: A private network that is physically secured to only the endpoints in the network.
A virtual private network replaces physical isolation of data links with cryptography.
If the network is not secured using cryptography, then it is not Private.
For example; in a MPLS Virtual Network; the Virtual network, would be subject to the possibility of intrusion into the network by the circuit provider, or any third party ---- there is no way to know that the network is actually Private then, so it is not.