When working on game reviews and first impressions for indie PC games, I often have need to ask the developer if there is an option I missed to allow key bindings and/or if they plan to add it into the game. Often, part of their response is along the lines of, "If you're having problems with the keyboard controls, you should try a controller. It's better that way."
The thing is, by the time I'm asking about things like key binding options, I've already tried using a controller. I might have liked it, might not have. Often I didn't, because for games that aren't action-oriented I generally prefer keyboard and mouse. When playing for myself, I also never play a first-person game with a controller if I have any other choice, and if I don't have a choice I often just drop the game rather than bother with it.
And to be honest, none of that stuff about the controller matters. I'm asking about keyboard bindings because it's a basic acessibility feature. Even for people who don't require key binding options as a prerequisite to being able to play a game at all, it's come to be expected by the computer gaming audience in general that games will offer the ability to customize the keyboard controls. Every time Total Biscuit does one of his popular WTF is...? videos, the first thing he does is go into the options and judge them, and key/controller binding options are always a thing he talks about. He does this because people care about it.
So many computer gamers prefer keyboard and mouse that neglecting to make the keyboard and mouse experience good in a game for PC is unwise. Do you want all kinds of people to play your game? You need to accomodate as many people as possible and key bindings help with that. Are you making a new entry in an established, turn-based genre traditionally found only on PCs for decades? You most definitely need it because your target audience will notice the lack and probably be very vocal in their reviews and other feedback. (Though if you're targetting one of those groups, you're probably a member yourself and fully aware of the need.)
There are a lot of indie developers who make a game with plans to expand from PC to mobile (or who are bringing a mobile game to PC) and try to design one common interface for both, but that's just a terrible idea. Many PC gamers expect to have the full capabilities of their computers used; this includes graphics and audio capability, but in addition to that they expect keyboard commands and the ability to rebind them, and if that expectation is ignored it will translate into lost sales and/or people just not spending much time with the game. Often, PC gamers will outright dismiss a game that feels like a mobile game to them, no matter how we might wish them to do otherwise.
I've been wracking my brain over the past couple of days trying to think of good examples of games that have good interfaces for both mobile and PC and especially ones that seem like they required a minimal amount of effort for developers to implement, but I really can't think of any. Spiderweb Software started with a game that can be controlled as mouse only or with a combination of mouse and keyboard and took that to tablets, adding a button to the equipment screen to compensate for the inability to do mouseovers. However, if I recall correctly, they had to rebuild the engine to make it happen. To be fair, the existing engine had been a work in progress since long before touchscreens were a thing, so it's reasonable to think that a developer starting now could build an engine with that in mind.
This blog post has turned into a bit of a ramble, hasn't it? The point I was initially trying to make is that good keyboard and mouse controls should be considered a must for any game releasing on PC. It really comes down to making the best use of whatever platform you're working with, though, and taking the time to get input and interface right for the people who are likely to be most interested in your game is of paramount importance.