This is not actually the case on the provider side.

The public Dynamic DNS is assigned by the provider belongs to a pool of IP addresses which can be assigned to any number of customers. The end user does not control the entire IP address block of the provider.

So, even when we add maps to our ISP interfacing device (like a router) to map externally assigned IP addresses to our internal addresses, we can assure the public address will remain static.

The issues of assigning static IP addresses on the customer side (the LAN side) is moot since these devices (and hence the private IP address blocks) are normally under the control of the end users.

For example, I have a fiber optic link directly from my ISP to my home. I do not pay for a static public IP address, so my address (on the public side of my router) changes all the time. I have all my private internal LAN addresses (192.168) configured static for a number of reasons.

It is not feasible to "punch though" from the public side to my private network because the public address changes constantly (as I have paid for, the cheap service) so we do not know what that IP address is from hour to hour. One moment, it could belong to my device, then an hour later the same IP address can be assigned to another customer of the same ISP.

I'm not following you about "production server" versus "personal server" and how this effects access using a publicly dynamic IP address. IP addresses do not care about their status. They are assigned based on a service contract with an ISP; and if you are paying for a public dynamic IP address, that is what you get on the public side, the fact if it is personal, experimental, fun and friendly, of the most important server in the world does not effect routing. The service we pay for effects routing and the configuration.

We pay for the services we need. In my case, if I needed to have (or wanted to expose) a server on my network accessible to the world, I would then pay for a statically assigned public IP address. I prefer not to expose my LAN to the world for security reasons.

On the internal LAN side, the point is mostly moot since we (the end user) have control over that IP address space and we can easily assign that as static (like I do) or dynamically, it is up to us and the size of our LAN users and how we want to manage things.