Here's the abstract from the Nicholls study.
Ball exit velocity (BEV) was measured from 17 experienced baseball hitters using wood and metal bats of similar length and mass but different moments of inertia. This research was conducted in response to safety issues for defensive players related to high BEV from metal baseball bats reported in the literature. Our purpose was to determine whether metal bats, with their lower swing moment of inertia, produce a higher linear bat tip velocity than wooden bats swung by the same players. Analysis using high-speed videography indicated significant differences in the x-component of velocity for both the proximal (metal = 5.4 m s~(-1); wood = 3.9 m s~(-1)) and distal ends of the bats (metal = 37.2 m s~(-1); wood = 35.2 m s~(-1)), p < 0.01. The orientation of the bats with respect to the horizontal plane was also significantly more "square" 0.005 s prior to impact (270°) for the metal (264.3°) compared with the wood bat (251.5°), p < 0.01. Mean BEV from metal bats (44.3 m s~(-1)) was higher than the 41 m s~(-1) velocity which corresponds to the minimum movement time for a pitcher to avoid a ball hit in his direction (Cassidy & Burton, 1989).
Note 41 m/s = 92 mph. Also note that Nicholls references the Cassidy and Burton study, and states that Cassidy indicates 400 msec of reaction time is necessary. The article Bob links also references Cassidy, but reads it's results as 325 msec. All other references I've seen to Cassidy refer to significantly higher reaction times than 325 msec.
Cassidy and Burton (1989) indicated 400 msec is required for a pitcher to complete a reactive movement to avoid being struck by the batted ball.
http://fulltext.ausport.gov.au/fulltext/2001/acsms/papers/NICH.pdfOur results suggest ball exit velocity from wood bats swung by live hitters is within, but at the upper limit of, human reaction time for defensive players. The finding that average exit velocity from metal bats was 43.98 m/s (98.95 mph), and as high as 120.97 mph, indicates a high potential for impact injury to fielding players.
The Nichols study that refers to a safe APRT of 0.425 seconds is here:
Nicholls RL, Elliot BC, Miller K, et al. Bat kinematics in baseball: implications for ball exit velocity and player safety. J Appl Biomech 2003;19:283–94.
And just to avoid confusion between the 400 msec number and the 420 msec number, they are referrring to the same thing. 400 msec is a calculated APRT based on ball exit speed, without taking deceleration of the ball into account.
However, if deceleration due to air resistance is factored in, the resulting APRT would be higher. Based on baseball deceleration calculations performed by Adair7 and substituting the physical properties of softballs, the deceleration due to air resistance in slowpitch softball can be calculated. If deceleration were taken into account for slowpitch softball, a softball would decelerate approximately 9.4% during the first 15.24 m of travel. Therefore the mean speed for a softball starting at 137.2 km/h would be 130.8 km/h—that is, 137.2 x (1 - (0.094/2)) km/h. Therefore the resultant APRT for an mean softball speed of 130.8 km/h would be 0.420 second.