because the metal used in shinano was worse than metal of 1930s from the report in US trials
The U.S. Navy Ballistic Limit (complete penetration minimum velocity with this projectile at normal) estimated at 1839 feet/second (560.5 m/sec), plus or minus 3%, which gives it about a relative plate quality of 0.839 compared to U.S. Class “A” armor (estimated, as no such super-thick plate was ever made in the U.S.). This was about the same as the best WWI-era British KC-type armor, which was what the Japanese were trying for - they had not attempted to make improved face-hardened armor, as the U.S. Navy did during the 1930’s, for actual ship installation.
The plate was excessively brittle internally, with too much “upper bainite” crystal structure due to too-slow cooling. This was due to using the same pre-WWI British Vickers KC-type armor-hardening techniques on plates over 17" (55.8cm) thick, for which they were never intended. This problem was solved during WWII, but no more VH was ever made except for some thin experimental plates. Brittleness did not seem to reduce resistance to penetration, though cracking might cause problems due to hits that ricocheted off.