the basic limitation of any empirical stat such as wRC is it approximates the influence of all of the inputs (through fractional coefficients, linear weights, etc). empirical stats by definition are approximations (ie. not 100% accurate)
BTS could have said the player is "50 +/- ?% better offensively than the league average hitter" with ? being the 2 standard deviations error of the approximation (2 SD is typically used)