Erm, even with his shell design, there was a big accuracy loss. From the Wiki (go there for original citations):
***
The gun designed to fire it had a 23,000 cm3 (1,400 cu in) chamber, a 45-calibre rifled barrel with 1/20 right hand twist fitted with a conventional muzzle brake.[2] Its breech was a conventional screw with interrupted thread.
Key performance data, from the Firing Table[1][3] are:
ERFB-BB shell, weight 48.0 kg (105.9 lb), M11 Zone 10 muzzle velocity 897 m/s (2,940 ft/s), QE 898 mils, time of flight 112 s, range 39.6 km (24.6 mi; 130,000 ft). Probable error in range 212 m (696 ft), in line 36 m (118 ft).
ERFB shell, weight 45.5 kg (100.4 lb), M11 Zone 10 muzzle velocity 897 m/s (2,940 ft/s), QE 881 mils, time of flight 99 s, range 29.9 km (18.6 mi; 98,000 ft). Probable error in range 189 metres (620 ft), in line 42 metres (138 ft).
HE M107 shell, weight 43 kg (95 lb), M119 Zone 8 muzzle velocity 675 m/s (2,210 ft/s), QE 764 mils, time of flight 65 s, range 17.8 km (11.1 mi; 58,000 ft). Probable error in range 59 m (194 ft), in line 12 m (39 ft).
The dispersion of the EFRB shell is more than three times that of the FH-70 field howitzer at its maximum range of only 5 km less, and is twice as great as FH-70s at 20 km (66,000 ft; 12 mi). Its maximum range with the M107 projectile is the same as any 39 calibre 155-mm gun and its dispersion about the same. (The “dispersion” figure means that 50% of shells will fall up to the stated distance either side of the mean point of impact, but 100% will fall within 4 times the probable error either side.) Dispersion of this magnitude significantly reduces the tactical value of the equipment.
***
It had *reduced* accuracy with his shell designs. It means your artillery is only effective against area targets and when used in close support with infantry you are likely to shell your own guys.
I have been fighting a one-man fight for a couple of decades against the "rangeophiles" at Fort Sill and MCCDC who seem to have forgotten that the old M107 175mm we had in Vietnam could reach out a wonderful 32,000 meters using Zone 3 but never, ever killed anybody on purpose. It had a Circle Error Probable (CEP) of about 1,000m at max range and a lousy frag pattern - you could sit at a surveyed target for the rest of your life and you'd be as safe as you would be, sitting behind the gun!
Remember the Kaiser Wilhelm Geschutze (Paris Gun)? Lots of Reichmarks, huge gun crew, unbelievable range, no combat usefulness at all.
Obviously, guided rounds have the potential of making long-tube artillery useful but they are very expensive and can be neutered with GPS jammers and plain old smoke to blank out laser designators. The only hope long-range gunnery has is to investigate further tube stiffness/shape migration sensing (tubes change shape while heating) and really great fire control algorithms combined with instantaneous MET. Might help, but it will take some serious experimental work to find out. Until then, anything that shoots past 25,000 meters or so is just an expensive noisemaker.