Ok. First, why do you speak of 25*25 when the tile size is only between 12*12 and 17*17?
That's the actual size, count the tiles. 12 is just on one side, 12 on another side, plus 1 in the middle, which is 25. 12 is used internally by the game, options menu should really say 25, so instead of 17 should be 35 (17*2+1).
Also, I'm no programmer, but since the games run SDL, wouldn't any more graphic thing end up being processed by the GPU? As such, since Cata's graphics are limited, you're unlikely to exceed your GPU capacity by increasing window size.
It's software rendering, so there is no use of GPU hardware acceleration. To make use of it program should run in full screen and use OpenGL, although some 2D functions might be hardware accelerated via SDL drawing functions I suppose. In any case it's not drawing bottleneck in this case, it's preparation for it, calculation of the interaction between the tiles and game variables that defines what and how will be drawn at the end.
Anyway, is it just me, or are you really averse to having people customize their game?
No, I'll help you make modifications you want.
AFAIK, changing tiles require a recompile (and since your code is hidden, no one but you can do it),
No, you can change or add new tiles by simply editing tile.bmp image, and in case of adding tiles also include the new entry in tileset.json file. It's really simple, look at "Graphics: sprites and tiles" thread for more details.
changing volume require messing with files,
Changing volume requires a click on the Windows volume control. Why do you expect such high polish of a free game in beta release? There were far serious and more important issues that needed sorting out first. When was I supposed to do all that? For what, for whom, why? Isn't it cool there is music and sound effects at all?
one can't change the window size...
Both terrain window and font size can be changed. The game is perfectly playable even with the smallest terrain window size people were playing with for the last three years with no one complaining about it. Yet you get bigger, and then suddenly it's not enough.
It seems to me you're making a game for yourself, suited to your particular need, with little regards to other people.
As if DDA team, or anyone for that matter, is making a free game according to what "people" want. You like all DDA changes? -- Anyway, the only thing you mentioned so far that actually has to do with preference is whether the character should be able to simply walk over bushes and such without requiring extra user action. I think if you weren't used to DDA you would think that's rubbish.
Moreover, one of the attraction of roguelike is large, rapidly evolving content, but here since nothing is in .json and the code is not public, the fact is that no one except you can add anything.
I find people actually prefer finished, balanced and polished games. I do anyway, and so my goal is to make a complete game where everything that is already in fits together, rather than having watered down bloat with never-ending balancing issues where playing feels more like bug-testing session. I am aiming for very focused strategic and tactical challenge, not time-waster where you can do many things with no particular goal. Besides, there is so much content in the game already, much more than many commercial top rated games have. It all just needs to be put together and make it all actually matter. There could already be dozens and dozens of different missions, each focusing on using different items and performing different actions. Making those missions interesting would provide much more gameplay value than hundreds of new items and weapons.