First I tune my basal until I believe it is about right: my BG doesn’t rise or fall all by itself when I am away from boluses, meals, exercise, and things like that. (Basal is the foundation, so tune it first.)
Then, if it is routine that after corrections my BG doesn’t drop as much as I calculated it should (even after 4 to 5 hours), that’s a hint that I needed more insulin, so strengthening my correction ratio would fix that. For example, I thought 1u of insulin should drop my BG by 25u, but if I routinely end up high, maybe 1u only drops me by 24u, or 23u. Keep adjusting until it’s working.
Similarly, if I typically find that after a calculated correction I end up going low, the correction was too strong, so I weaken the correction ratio to fix it.
If it is routine that after a meal bolus, my BG doesn’t come back down to about the same place as where it started (even after 4 to 5 hours), that’s a hint that I needed more insulin, so strengthening my insulin:carb ratio would fix that. On the other hand, if I find myself routinely going low after meals, I weaken the carb ratio so I don’t continue to get too much meal insulin.
Try this experiment. Fill a syringe with 1u of water (why waste insulin). Don’t inject it, just deliver it onto your skin to see how big a drop you get. How does that compare with the amount of leakage you are seeing from insulin injections? I’m guessing that your “lost insulin” from an injection is much less than 1u, in which case you wouldn’t want to take a whole extra unit when you inject.
Back when I was on MDI, the syringes I had were 30u with 1u markings, but I routinely dosed to the nearest 1/4 unit by filling the syringe with the plunger stopping between the lines part way between the marks if I wanted a fractional amount more. There’s no rule that says you’re only allowed to stop filling the syringe when the plunger is exactly on one of the lines. Of course if you are using a pen you don’t have that ability to choose in-between amounts.
My point is that if, after trying the water experiment you think that you are losing 1/4u of insulin every time you inject, I suppose you could increase your dose by 1/4u every time, but I’m guessing the leakage is so small it doesn’t really matter.