The way I view it is that modders are getting license free access to the engine, free modding tools, the game libraries, publishing of the base game, etc.
It's completely unique. You can't make a film or a song and publish it using unlicensed stock footage or samples and not have it be cease and desisted even if you give it away for free. That anyone can use Skyrim, for example, as a free platform for their content is not something that has an analog in any other field, and it is something that games companies spent time and effort in supporting. It's clearly a burden or a risk for devs, because there are so few games that support modders to such a degree.
I get that modding is more than "practice" for a lot of people, but if it's getting to the point for a modder that they're spending so much time and money that it's not worth it, they really should either make the mod less ambitious or they should take the next step and start developing their own games and take the risks to reap the rewards.
There was a time in the 2000s where you couldn't make a game in your garage, not like you could in the 80s or 90s. But now, with such robust tools and favorable engine licensing agreements, there are plenty of self-published games made by one, or a handful of people.
I could understand the argument if it were still like the 2000s where realistically, the only way to make a game would be a free Half-Life or Quake total-conversion. For those devs, the fact they couldn't make a stand-alone game was because practically speaking, they had to make a mod. Tools were too expensive, too undocumented, and too demanding for a small team to create a game from scratch -it was just not realistic.
Today modding does not have to be the primary road to developing a complete game experience in an affordable way with a small team. Modding is a conscious choice for small, or inexperienced teams so they can reduce their risk, but they accept that their reward is also reduced.