It's really your decision, how many details do you want to know about game development. The only thing you need to consider is that although modding is much faster than programming, it probably wont teach you how to make your own games, while by learning to program them on your own, you will also learn how to alter already existing code. Therefore, I'd say that it would be better to learn how to program your games from scratch (if you have the time and patience to learn it).
This is just my opinion, good luck with whichever option you go with.

i think for some people modding may be a gentler introduction to design.

Yes, but I doubt that they will get far - modding is a way to practically immediately see the effects of your code in a "distributable" game. Let's put it this way: you can perceive game programmer as a fully devoted developer, and modding as a hobby.

Hi daedalus, I have been learning 3d modelling and graphics for the games industry, I got started because of modding, and when I found the course I'm doing now and realised how close some of it is to what I was doing I signed up. But as I've learnt more and gain knowledge, I've come to realise that modding a game, isn't anything like making a game from scratch.

As you mentioned, when modding a game you'll only scratch the coding and scripting, for older games this is just opening a script in notepad and copying and pasting things around, new games with kits wont even touch that. When I first saw python and other real program systems I was totally lost in the jumble of words, still am, and most mods are simple texture recreations, making new levels or characters, and many new assets. So for the most part depending on the game, it just copy and paste or replace a previous asset.

Complexity is also a factor. For example Left for Dead 2, there are a lot of new levels, No new weapons or power ups to my knowledge, so the emphases is to make a good challenging level, through it's level design, and some of the new level are better than the stock levels. But that's all,, On the other hand,, an old space game called Freelancer by Microsoft. A team has spent years making a mod called crossfire, which won an award, they found old cut scenes that were deleted created over 100 new levels, ships, weapons, clans and dockable planets and outposts. They messed with the AI, and recreated the UI and load screen, basically in every way they've completely retrofitted the game. My point being making a new level for a game can be a challenge, but completely rebooting a games means knowing it inside and out. The engine and all the coding. Also your right the modding kits that were made for oblivion and fallout 3 make the whole process a lot easier. So there is a lot less to learn and explore.

From my perspective, modding was a good start, I had to learn some coding, graphics and how the game worked. Also there are always websites with tutorial, as well as open GL programs to accomplish the tasks. It gave me something to aim for and it was a lot of fun personalising my favourite games,, but it certainly didn't prepare me for what is really involved in the industry.

Telling people to do something is easy. Telling them what to do is harder. Telling them to do the right thing is difficult. Having them do the right thing after you've told them to do it is challenging at best, maddening at worst.

Management on the surface is easy, but peek under the vaneer, and you realize its a massive juggling game of knowledge, respect, schedules, risks, etc. You cannot expect to manage a technical team well without having some moderate experience in the technical field and you can never expect to manage a team well unless that team trusts you. period. Sure, you can slide by on intimidation or ignorance for a time, but it eventually catches up either in the project failing or you being fired. Good game designers imho have the experience and learned lessons from "having done it", not from "having watched someone do it".