1) It means the authors did not run any tests outside of the published limits, so your results could be applicable, or they could be very wrong.

2) Same answer as for #1.

3) This is where engineering judgement and experience come into play. If we made this an error, then a Dn/Tn ratio of 19.999 would halt the analysis. Would you really want that?