Of course, with more servers, a VPN service can more widely distribute demand from consumers, giving each customer greater speeds.
However, ignoring this particular effect, does a VPN service see speed improvements given more servers?
I was told by someone with experience in Azure enterprise applications (who was discouraging me from buying a VPN service with relatively few servers – to me, it seemed to have compensated with its smaller user base – its real-world performance was solid) that, given more servers, a VPN service will perform better.
His rationale was that, in a manner more or less identical to an virtual enterprise WAN, the VPN servers rented by the VPN provider (something analogous AWS I suppose) would have dedicated high bandwidth connections between each other as a result of optimizations offered by the hosting provider. These high-bandwidth connections are supposedly just parts of the open internet. This is different from the generally high internet speeds offered by the server provider. In a virtual enterprise WAN, this allows for fast communication between different servers in the enterprise setup. However, for a VPN, I can’t really understand why this would be relevant.
He says that, instead of a VPN server directly accessing the open internet to query the target server (i.e., the server I as an end user wish to access), the VPN server uses these dedicated high bandwidth connections to communicate with another VPN server that is closer to the target server. This supposedly eliminates slower and/or an increased number of nodes to reach the server.
I lack experience in this altogether, which is why I wanted to ask here to understand if something like this actually has an effect.
Insight would be greatly appreciated. Thank you!