I know I don't post often and this is certainly a strange post after several months, but I think I can help on some of this.
Viruses did not necessarily have to be so diverse and widespread before the fall of man. There could have just been a few types, all of which had a useful function. Since viruses require other cells to replicate, they are technically not even living organisms, although that is a human classification. They are at least unlike any other living thing ever created. I do like the bacterial population theory.
However, regarding the transport of good DNA, it is important to understand how a virus works. Viruses, even if they carry a strand of DNA or RNA, must invade a host cell, inject the genetic material contained within the envelope and hijack the cell's processes. The cell then begins making many, many copies of the Virus until the cell wall ruptures, killing the cell. This rapid replication gives PLENTY of opportunities for the genes to change. Any slight change in the genetic information will change the configuration of the virus, making it more or less effective against the host. Some viruses change so rapidly (the flu and coronaviruses are good examples) that a body has difficulty producing antibodies for each strain that appears. That's why you get the flu every couple of years. The virus has changed and the old antibodies your body made are no longer effective markers for the updated version of the virus.
Now, you need to understand what the genetic material is for. The genetic information contained within a virus is basically instructions on how to build the virus envelope, or the shell that contains the genetic material. It works similarly in your own cells. Each cell contains DNA that codes for making more copies of itself. So, a virus must contain DNA that codes for its own self. Let's say someone has a genetic defect that causes them to have a protein deficiency. If we had a bunch of empty virus envelopes and put a strand of DNA in them that coded for the deficient protein missing from the person, the virus would attach to a cell, inject the DNA and the cell would begin to produce that deficient protein. However, it would not make more copies of itself. You would have to continue to inject these viruses because they would be unable to replicate since the contained DNA actually creates something different from the virus itself. This is a VERY expensive process. And eventually, your body will create antibodies to the foreign virus and the treatment will be rendered ineffective.
If you wanted, somehow, to make a self-replicating virus that creates proteins that are needed in a person with a genetic problem, you create a terrible conundrum. The virus will act exactly as it always has--it will replicate and kill cells. This replication process, although also coding for proteins that are needed, will have many, many opportunities to mutate. If this mutation occurs, it may not only change the virus, making it more dangerous to a person, but it may also modify the protein, which would render the more virulent strain completely ineffective in helping supply deficient proteins.
It would be like handing the reigns to a hamburger restaurant over to a bunch of kids. They might start making hamburgers, but after a while, they are going to be making whatever else they want, and not very effectively. They'll probably wind up burning the whole place down, to be honest.
Hope some of that info is useful.