I don't know how profitable the games industry is, but I don't think you can just look at examples like Minecraft - these are the rare exceptions, not the rule. Of more interest would be things like median averages of things like sales, revenue, profit. The revenue numbers of the games industry for a whole are impressive, but I don't know what things look like for a typical indie developer.
In terms of not needing to advertise - yes some people can know about them, but it's a question of numbers. In some cases it can lead to large numbers, but again those are the exceptions. I see from my own set of open source applications the range that is possible - the most successful (not a game) gets more downloads per day than all my others get in a year.
There is the question of expectations - even for my games, I can still say I've had thousands of people try them, and hopefully at least some of them like it (and from reviews, emails, I know that some of them do). So in that sense I've both achieved my goal of having fun writing a game I wanted to write/play, and had other people enjoy it too. Making enough money to make a living would require a lot more (of course if I was working full time on it, I'd have a lot more time to invest; but it'd be a risky business venture for me to take).
Having said that, it's not clear if that's the reason for Open Source game development (or alleged lack of). I don't think Open Source developers in general do it for the money. Regarding the Reddit quote "the skill/talent gap is just too large" - the biggest factor is time. In the commercial world, this also means money. For Open Source, this has typically come from people working on projects in their spare time.
Some non-game Open Source projects do benefit from funding from companies, I don't know how much of a factor that is.
The usual answer for a difference that games have is they require collaboration from both developers and artists. I think even if there are just as many artists, the fact that this collaboration between the two sets of people is an issue. Collaboration is in general harder (when people are working for free in their spare time, and aren't in the same office); many famous projects (Open Source or commercial) started from a single developer, which is therefore harder with games.
I think another factor is not that there's a lack of artists as such, but there may be a lack of people interested in working on Open Source projects? Arguably the "Free / Open" philosophy doesn't seem as common among artists (after all, the movements came from software developers, and I'm not sure the art world has had any similar kind of movement - Creative Commons is the closest attempt).
For the clones or free engine implementations of commercial titles, I consider them a waste of time, since 1. no original new content is created, 2. it is still not open source
Open Source engines are a good thing to have, no different to it being a good thing to have Open Source development tools of any other kind.
They might not end up being used much or at all - but like games, any software may end up being popular or not.