The critical L of Twaithes is L=-0.09 obtained from a fit of data obtained numerically before 1948. Comparison of this theory with the 3th order Polhausen method indicates one could use L=-.099. The 4th order method yield a larger |L| and is known to be quite inaccurate when compared to experiments. In our experiences the value of L=-0.0992 does better compare with experiments than the value used by Twaithes.
(There is a typo in the original paper: Pelorson X. et al., J. Acoust. Soc. Am. 96 (1994) 3416-3431. Please consult for correction: Hofmans G. et al., J. Acoust. Soc. Am. 113 (2003) 1658-1675)
Also there is no justification to assume a continuity of the second derivative of the velocity as needed when using the 4th order Polhausen method.