At about 40,000 yards, the U.S. Navy 16"/50 firing a 16" Mark 8 Mod 6 AP projectile (the later Mod 7 and Mod 8 designs were post-WWII, so I usually do not count them and they were no better ballistically, to my knowledge) will hit at about 45° downward angle and 1607 feet/second (489.8 m/sec). Just as with a point blank hit at 2500 feet/second (762 m/sec) and 45° obliquity, this hit too will barely hole the plate as the projectile is hitting at 0° (normal) obliquity, though not completely penetrate it. Any slight barrel wear will lower the muzzle and striking velocities and no holing will occur at these or any other ranges, as mentioned. However, this is so far above any real fighting range (even with radar it is hard to see the target due to the earth's curvature interfering, especially in any kind of imperfect seeing conditions) that I do not even consider it in my computations, while putting the gun barrel up to almost touching the enemy turret is also a pipe dream in real life! Thus, no holing or complete penetrations, ever, though possibly some cracking of the plate and possible jamming of the turret if the crack-off plate piece is dislodged badly enough.
Therefore, these plates are the only warship armor plates that could not be completely penetrated by any gun ever put on a warship when installed leaning back at 45°, as they were in the actual turrets!!! Even to completely hole the plate all the way through at that inclination requires a brand new 16"/50 Mark 7 or German 38cm SK C/34 gun at point-blank range firing the latest versions of their respective AP projectiles; it might be cracked at a lower striking velocity, but no hole put entirely through it! And they said guns had completely overmatched all armor - not so!!!"
Well, there you have, that's how the tests (2) were done. Very idealistic conditions.