Sometimes culture prevents/subverts this. For example, if your reviewers won't approve diffs where the new code is not yet actually connected to anything, you w

Sometimes culture prevents/subverts this. For example, if your reviewers won't approve diffs where the new code is not yet actually connected to anything, you won't be able to parcel a feature into small diffs and then turn it on at the end. That is only one example. Also, while it's true that reviewers will "lgtm" a large diff without close examination, reviewers also allow themselves to get hung up on mind-bogglingly dull details in a small diff. There is little else to focus on in a small diff, and people seem to feel insecure about "rubber-stamping" PRs. There apparently needs to be some contribution from them, or else do they really exist? I usually just change whatever silly thing they want changed, so we can all get on with our lives. And because I strongly believe we shouldn't get too attached to our own code or ideas anyway.