The original note, as written at the time
In this concept, there must be two portals to achieve teleportation between two places. So efforts must be made to position the second portal at the destination as early as possible. Say we wanted to send an expeditionary team to Proxima Centauri. We would need to find a way of creating the portal at the destination at light speed. This would be critical for the success of the process. Even then, we would have to wait four years.
Ongoing development
Teleportation between two places requires two portals, which means the hard problem is not the technology but the seeding — positioning the second portal at the destination before you need it.
Send an expedition to Proxima Centauri and the arithmetic is brutal. The second portal must be created at the destination, and it can only be created at the speed of light, so the wait is four years before the first person can step through. The portal does not beat the distance. It only beats the second journey.
That reframing changes what the technology is for. A portal network is not a transport system, it is an infrastructure investment with a very long lead time — closer to a railway than to a car. And infrastructure with a four-year lead is only worth building if you intend to use it for a century.
It also means the first thing through cannot be a person. It must be a machine that can build the far-side equipment, which is a different engineering problem and arguably the real one. Build a self-replicating factory that can survive four years of vacuum and then assemble a portal, and the transport question answers itself.
The strategic consequence is uncomfortable and worth stating. Whoever seeds the portal owns the route, and a route that takes four years to open is a route nobody can contest in the meantime. This is the most colonial technology imaginable.
So the design question is not how fast the portal goes. It is who gets to place the first one.