I would merge for simplicity and better looks, accepting slight mismatches that are purely visual.
But I ignore any backgrounds. I want a single color that is distinct from any terrain (thus should be black in 95 % of tilesets). I see that enough people care about backgrounds and that there is a problem that I don't experience.
Could it be that sometime in the future, the Sonic tilesets I made will get HD makeovers and then people will prefer those over the originals I created, it just feels like those who cannot make tilesets up to the standards GigaLem makes them are going to get written out effectively
Let's assume that your scenario becomes true, and other people improve your tilesets. The reaction should be the complete opposite of "
":
1. Your sets got used, thus you already have enriched culture.
2. People care about your sets so much that they improve on any few remaining bugs. Not every set gets such honor.
3. Pepole strive to be 100 % physics compatible with your work. Wow!
You are not written out when people improve on your work. You opened the door for others to join. Be emotionally tied to your work, not to the results of your work.
I want to make sure it looks good with the tiles I have available, and take care to do my best to see that no pieces stick out awkwardly; such that large chunks of terrain made of several tiles mesh well and look like one large piece. Changing how the tiles look can alter this balance.
Yes, this is the most important argument against the merge.
I'm mostly concerned about the precedent that any replacement would set, rather than against this specific replacement.
This is slippery slope and should not count as a counterargument.
official support for re-skin tilesets is ever implemented.
Nobody will bother about re-skinning levels on the client side. The exception is disabling backgrounds because backgrounds distract. Stuff should be as good as possible out of the box.
In this light, I like how you will adapt your own levels to Giga's sets.
-- Simon