USN's 16"/50 barrels did have droop. It was corrected for inside the MK 8 Rangekeeper with a simple cosine cam and whatever droop there was got put in the elevation gun order section of that fabulous mechanical computer. But the flat ring the turret circled on, called a roller path, was never hand scraped. It was turned as good as could be then set on the deck and welded in place. Once the turret was set on it, it stayed there forever; unless the ship capsized then it would just drop out and fall to the bottom as only gravity kept it in place. The gunners mate's could get to it, clean and lube it so it worked well and lasted a long time.
Note the turrets had 18-inch thick faces and a bit less on their tops and sides. If you look at one, their back end sticks way out behind the roller bearings on the round platform they move on. The weight of the barrels is well balanced. But the plane of each of the three turrets "roller paths" wasn't parallel with each other. Unless this was corrected for, each turret's guns would have a different elevation angle for a given elevation gun order. This problem was solved with a "deck tilt correction" adjustable gear box in each turret. A fire control technician (USN abbreviation FT, whom I was for 22 years in the USN) in the main battery director would put its sights on the horizon, the director signals for bearing and elevation would be sent to the rangekeeper in main battery plot where the FT operating it would set the gun order correction section to zero. The turret barrels would then point where the director pointed. Everyone was on sound powered telephones and when the FT in the director sounded "Mark" as the director sight crosshairs crossed the horizon on the up roll, the FT in the turret looking through a bore scope in the breech would see where it was. If it wasn't dead on when he heard the "Mark" he had to adjust the deck tilt corrector a bit. This was repeated every 10 degrees of the turret's firing arc, and on each turret, too. After 'twas all done, each turret would have a chart with its high path point and each 10 degrees correction needed put in the deck tilt corrector. Looking at these charts made it easy to see how each roller path had its own high point and amount of correction needed. This "bore sighting" process took a while and we all prayed for calm weather so it would be easy to do. And it was checked and corrected for a few times each year 'cause big ships bend a bit from use and heavy seas. 8" and 6" turrets as well as 5" gun mounts all had to do this.
The rangekeeper also had a parallax corrector for each turret on board. As the turrets center of rotation was different distances in yards from the director, each one needed different corrections.
Barrel life, finally, was rather short. Considering a full charge for a 16"50 was six 110-pound bags of powder; a reduced load had 4 bags, there was a lot of wear at the breech end. Each with its own packet of black powder on its rear end to enhance burning. In their first years starting in the late 1930's, barrel life was about 250 rounds before the throat had eroded to its accuracy limits. A change in the powder in the early 1940's increased it up to 350 rounds. But after each 50 to 75 rounds, the liner with the rifling in it would be pushed forward enough that it's muzzle stuck out beyond the rest of the barrel. When its protrusion was up to 1/4th inch, it got turned off back flush with the muzzle. A special hand turned tool was clamped on the muzzle to turn off the extended liner back to the rest of the barrel. After refacing that liner 4 times, it would be replaced when it stuck out about 1/4th inch. It took two days of hard work by several people using large machines to pop that liner out then replace it with a new one. The old liner came out the back end of the turret.
FT's also had computers to calculate muzzle velocity using the powder lot's tempurature curves A bore erosion gage was put in the breech and the reading compared with known wear curves vs velocity loss. This was all added together and the muzzle velocity settings were cranked into that mechanical rangekeeper.
Not a bad way at all to set stuff right. The battleship New Jersey's Fire Control Smooth Log (detailed historic data of how the guns performed) I got to see in 1968 listed their first-round miss distance off the North Vietnamese coast at 90 meters. Pretty good accuracy for those big guns