The really hard part sometimes is getting political support within an organization. Someone usually went to a lot of trouble to argue that a feature was worth building, and a lot of senior people had to agree to it. A bunch of folks worked on it. Now you're asking them to admit that it didn't get traction, and more work should be done to remove it (and maybe annoy some customers along the way.)<p>It's necessary, and it should be done, but it's often unpalatable and doesn't happen, leading to bloatware.
As the question is posed, it's not clear to me if the feature is in a product that customers are using, or if this is a feature that just didn't work out and never was in a distributed product.<p>The two are quite different. Killing a feature customers are actually using can create problems. But in some cases not killing it enforces the wrong architecture on the program.
It's an interesting question but I think the main focus has got to be that you're trimming fat. I'm sure that all the effort produced still provides some valuable lessons that will save you time in the future. Focus on the positive and keep moving forward. Don't be a hoarder.