I think this has more truth to it than most here are going to be willing to admit. Nothing wrong at all with puzzles, but when you've got no AI, no collision detection, no complex models to build and animate, and so on, well, that certainly makes it a little easier to get the game done, doesn't it?
As an indie developer... absolutely.
AI, collision etc aren't really issues, though. The real issue is the pipeline for character models, which goes something like...
Modelling [+retopology] -> rigging -> animating [mocap, cleanup, animator]
Also in the pie, texturing and voice.
And combat-related means a lot of QA, and a truckton of vfx\sfx to get through. And environment to place those in.
Puzzles are cheaper to develop. On the other hand, you could swap the question around and ask why the average AAA game has a lot more combat than most smaller titles on the market, because they wouldn't just choose it because it's more complex.
No reason to assume that combat should be the norm and puzzles the exception, outside of audience expectation. I think most AAA devs think their audience is scared of anything other than constant progress, which is why puzzles in Uncharted 3 had the option to solve themselves, and ancient dungeon puzzles in Skyrim required the skill level for pattern recognition of a small child.
I think it all comes back to playtesting, where a player is more likely to give up on a puzzle if they can't work out the logic behind it- we've all run up against stuff where, once you understand the puzzle, only then can you start to solve it. Give a player a tough combat encounter and they already understand the basic rules and the tools in play, so are more likely to give it one more go, especially if they do a bit more damage/get a bit further/last a bit longer each time. It's like running up against a stone wall compared to running up against a tough slope- one looks doable and the other appears like a dead end.
To be fair, a properly formulated complex combat encounter can be seen as a puzzle. A time-sensitive, semi-randomized one, but still. Ask XCOM players about that : )
One legitimate observation, though, is that indie game developers are absolutely more comfortable with delivering much more challenging products. Puzzles only make sense if they're challenging, and that means puzzles with a reason to exist are over the challenge cutoff for AAA productions.