There are two trends imo. Some titles keep using Flash especially if it's kind of huge project relying on desktop users. And there's slow transition from Flash to HTML5 (and also mobile using wrappers) for small games.

And for "browser games" they are also in transition with HTML5 (like recent HTML5 Travian) as market is now more demanding than mere "Excel sheet wrapped in cute graphics"

It's not dead, just not it's golden days.

Mobile games are more popular than web games because developing them is very similar but the income from web games is minimal by comparison.

there is currently no incentives for web technology to improve. Infact there are many reasons the big firms (apple, google, microsoft) dont want to see any form of desktop/native experience in a web browser.

1) They want you to use their proprietory technology to build an "app" as you have to pay for this and you'll get locked into their dev system which encourages you to continue making apps for their platform.

2) They want to lock their customers into their eco system (though not as much a problem as it was the feeling amongst users is still "i cant switch to PlatformX because i have all these apps on PlatformY")

3) They want to sell apps through the app store and not allow apps to be accessed through the web. Only text/video is encouraged here.

and the dev tools / code are different between each platform. Fab...

Of course this is the exact opposite of what all of those people championing HTML5 wanted, we were told canvas and opengl would deliver games.

We are having the wool pulled over our eyes, W3C is a total farse.

Well, I am working on a browser game, I've never been fan of "small Flash games" but there is a market for them as well where people earn money mostly from licensing their games to networks. ( https://www.truevalhalla.com )

But still I think mobile games are the worst where CPA is high, competition is fierce and it takes sheer luck or serious money flow to be noticed.

And better don't have very high expectations from Canvas or webGL although they are not very bad, just not suited for every task.

How much on average does a popular web game make?

Hard to say. However, it appears that the entire web game sector accounts for just ~5% of game industry revenue, vs ~25% for mobile.

I was about to reply $0, but then noticed you used the word "popular".

The vast majority make nothing. The median value is $0 profit.

But if you only take the popular games, which means you are only taking the outliers, they can make quite a lot of money. I don't think anyone knows the average probably because it is meaningless. The average net worth of people in a bar may be tens of thousands each, until Bill Gates or Jeff Bezos walks in, then on average everybody is a billionaire.

Many sites will be making dollars, some will make hundreds of dollars, some will make thousands of dollars, but when you throw in a site making a fortune their average appears considerably higher than their median.

Statistically YOUR game -- if you ever complete it -- will not be profitable. However, if you iterate over the product with a strong business plan which begins before the product is even designed, you can almost certainly beat those statistics and make a profit, since most web games don't have any business planning.

Of course this is the exact opposite of what all of those people championing HTML5 wanted, we were told canvas and opengl would deliver games.

I feel like there's this weird sense of WebGL + Canvas + HTML5 not being capable of delivering games, but I'd argue that's not true anymore. WebGL support has grown enough in the last while that I'd feel confident using it to make some pretty decent games- and there's a reasonable amount of stuff done with Unity or Unreal's web players (or Phaser etc. or from scratch) to show that it's doable. There's a very low barrier to entry, and you get multi-platform out of the box (given your game gets along with different browsers). The real issue, IMO, is that there's not really any money in it. If the goal is for the game to be a direct source of income, then web isn't really a viable platform - but advertising games, kids games, content associated with other media or something, stuff to drive traffic to your website, etc. make a lot more sense.

I think we see fewer web games in the sense of what you'd see on newgrounds mostly because mobile has hijacked that market, and because indie devs have a lot more options now.

Of course this is the exact opposite of what all of those people championing HTML5 wanted, we were told canvas and opengl would deliver games.

I feel like there's this weird sense of WebGL + Canvas + HTML5 not being capable of delivering games, but I'd argue that's not true anymore. WebGL support has grown enough in the last while that I'd feel confident using it to make some pretty decent games- and there's a reasonable amount of stuff done with Unity or Unreal's web players (or Phaser etc. or from scratch) to show that it's doable. There's a very low barrier to entry, and you get multi-platform out of the box (given your game gets along with different browsers). The real issue, IMO, is that there's not really any money in it. If the goal is for the game to be a direct source of income, then web isn't really a viable platform - but advertising games, kids games, content associated with other media or something, stuff to drive traffic to your website, etc. make a lot more sense.

I think we see fewer web games in the sense of what you'd see on newgrounds mostly because mobile has hijacked that market, and because indie devs have a lot more options now.

Could you make, like $300 bucks off web games?