Why is Ballgown still using FPKM as normalization method?

For some time (5 years), we know that FPKM is the (maybe) the worst normalization method ... Maza et al., 2013 (https://doi.org/10.4161/cib.25849) said that we should stop to use that ... "it clearly emerges that the FPKM normalization method should be avoided in DE analysis" so, WHY ARE WE USING FPKM ??

The normalization method is (almost) the most important step in DGE analysis and if use the worst method, are our result trustworthy? Should we go throw the last step of "new tuxedo" protocol away?