Rejecting an improvement because it's just "too much code" is... brilliant, disciplined, and extremely difficult to do. Such a high percentage of software development is maintenance, and it's easy to underestimate that long-term maintenance cost.