Shovel Knight has been partially done by TerraEsperZ, no idea if they're still working on that
Unfortunately not. I've always mapped using screenshots and, whenever possible, emulators that could disable layers. Shovel Knight has many layers that can't be disabled so you're left with two choices: take a ton of screenshots and extrapolate what you can't see using the game's individual tiles (a gruelling task), or extract the layers straight from the game's code. I'm nowhere near smart enough to do the later, and trying the former took me several weeks to assemble the game's simplest maps.
As much as I love this game, I simply don't have the patience and energy to keep mapping it. Sorry 
.
The extrapolation is what I tend to do, but yeah, it's really not fun if you can't get lost in it, and when it's too many layers
and the layers all share colours, it just gets frustrating.
Since you're not working on it, I'll keep an eye for it in giveaways and the like. Used up my vidya budget on Momodora, but I'm curious to try to map SK too.
@Nega: Ripping models from games isn't easy. Even ripping assets from classic games is hard (despite some emulators having tools for it!), which is why so many maps are made using screenshots. It's easier for some games than others, but there's no one-size-fits-all method, because different games use different methods for storing their assets. A tool developed for one game isn't likely to work for another game, even if the developers used the same game engine and the same 3D programs. In addition, it's a lot easier to make a case for sharing a finished map with the public compared to sharing raw assets, especially when it comes to modern games.
If you're looking to commission people to map these games for you, you should first sort the games into categories based on what types of assets they use, so that you can approach the right people for each game, and can have a general idea of the amount of work that would go into it. Mapping a game can take dozens, sometimes hundreds of hours, and you'll want to budget accordingly, and perhaps focus on the games that are easier/quicker to map.
2D pixel art games
- These are probably the easiest to map, even if no tools are available to disable parallax layers
- If tile-based, it's easier to reconstruct parts that can't be seen during normal play
- If the game has parallax effects, that makes it more difficult. If the different layers can't be disabled to allow them to be captured separately, it requires a LOT of screenshots to get the image data needed to look nice on the map, so be prepared to pay for many extra hours of work. If the game uses the same colours for the different layers, separating them from screenshots is even harder.
- If the game has dynamic lighting effects or colour overlays and no way to disable them, again, be prepared to either pay for a lot of extra screenshots and manual recolouring, or expect a map with choppy colours, especially of the lighting effects consist of gradients rather than sharp shapes.
2D non-pixel art games
- If the assets have aliasing (soft, non-pixelly edges), it's much harder to rip them from screenshots. This is not a problem if the assets are static relative to each other, but is an issue if there are parallax effects, as the assets will be overlapping different things in different screenshots, which will make them look slightly different in each one. The softer the edges, the worse they'll look.
- As with pixel art games, lighting effects pose a challenge for screenshot mapping. Manual recolouring is generally not an option with non-pixel art because the individual pixel colours aren't as predictable and there are way more pixels to work with. In those cases, one needs to either rip the assets from the game and rebuild the map from them, or one needs debug tools to disable the effects or fix them in place.
- These games tend to have high-res art, which means that to be reasonably-sized, the maps will likely be much smaller than what you'd see in-game. Plus, unless the mapper is using ripped assets, they'll be dependent on the resolution they ran the game in.
3D games
- If you want maps that represent the in-game visuals, you pretty much need access to a level editor where you can control the camera or some very good asset/level extraction tools. Even with such tools, 2D maps are only possible for games where the gameplay areas are either essentially 2D (e.g. Trine), or are flat enough that they can be separated into a series of flat "floors"; any game with a more complex use of the 3D space will generally not make for a good 2D image.
- If you want maps using in-game assets, it may help to sort your desired games by the engine they use. Some engines support reverse-engineering better than others.
- Most people just screenshot the in-game maps and annotate them, since the developers would have already gone through the trouble of figuring out how to make their 3D space work in 2D. If that's not a good option, drawing a map from scratch might be a possibility.
Another thought:
Sometimes, engine publishers will make parts of games or whole games available as project files (as opposed to compiled games ready to play), in order to promote their engine. YoYo Games did this with Spelunky and a few other games a while back. Tracking those down could be quite helpful to mappers. In general, finding resources and tools to help mappers with specific games is probably even more likely to get those games mapped than offering money. Money won't get a game mapped if the mapper doesn't have the tools to do it.
Since Bastion's on your list and I have it, I tried to see if I could rip its textures. Nope :| Couldn't get any of the tools to convert the XNB files to PNG files to work, they all crashed. Hopefully there's someone else out there with an interest in this game, someone who's more familiar with XNA, its files, and its tools. Bastion's maps are, as far as I can tell, made of tiles, so without a way to open the map data in addition to the getting the textures, I'd have to arrange the maps out of tiles by hand based on screenshots, and that's too tedious and time-consuming to be worth it.
