Optimal fuel droplet size has been shown to make a few more hp with both carbs and EFI. The most I've seen came from dyno tests from a friend who has built multiple SCCA championship winning Nissans. He saw just over 2% by going to different injectors and higher fuel pressure. Too small on droplet size and you lose hp, too large also. I might expect lower gains on a slow revving engine like a Lycoming but don't know that for sure. The droplet size affects combustion speed and how heat is released. You'll see many SAE papers on this, mainly to do with emissions.
On 540 engines with vertical induction the EFI throttle bodies have lower pressure drop than the carb or RSA-5 servos, this allows more airflow, hence more power.
The RSA-5 servo doesn't present much restriction to a 360 though so gains are lower there. There appears to be no drop in MAP on 360s with our std 60mm TB. You can see what you have now for MAP, fly to sea level and go to WOT. You may have some extra pressure from prop inflow but if you only get 29 inches, that would be about 3% loss. The carb will usually show somewhat higher loss than this due to the venturi restriction.
On 540s with horizontal induction and the RSA-10 servo, pressure drop is minimal so I'd expect to see even less hp gain there with EFI. We use an 80mm TB on this application. Anything past about 2/3rd to 3/4 throttle shows no gain in MAP.
Realistically on a 360, I'd expect to see no more than about 6hp gain with EFI over a good RSA-5 or AFP setup with matched nozzles.
If someone can produce a back to back dyno test on the same engine with mechanical injection and then EFI, those would be interesting results.
I'm doubtful you'd see 15 hp on a 360 but welcome to be proved wrong.
When testing engines or many others things, you should use the scientific method- change only one variable, repeat and observe. Too many people will build a new engine, port the heads, change the exhaust, raise the compression, change the manifold and add EFI, then attribute all the power increases to the EFI. Not realistic conclusions.
Of interest, when we ran a Rotax 912 on our test stand with a club (known power absorption curve), we saw a 3-5 hp gain with the EFI. This was using the original Bing carbs as throttle bodies for expediency but it also removed any variable using different throttle bodies. The 912 is known to have unequal fuel distribution with the carbs. The EFI, even without individual fuel trim in that era, showed a much closer EGT grouping at WOT compared to the carbs. We had a wideband on the engine and tried to bracket the EFI AFRs where the carbs had been running.