If you do a task daily, and you can shave one second off it, you save 30 minutes in 5 years. So as long as you spend less than that developing whatever tool or technique making it more efficient, you end up saving more time than you spent optimizing.
I love this and am glad to see it again after all these years. Other considerations, in addition to time, might be 1) whether the project taught you something new, and 2) if there is some mental or physical health benefit to automating the task