Neither element set format encodes the century of the element set, so I suspect that whatever math is being done in the program to determine the age of the element set will fail in either case.
You can test this hypothesis by making up a verbose element set manually with your favorite text editor.
Decode two line element sets with the following key: 1 AAAAAU 00 0 0 BBBBB.BBBBBBBB .CCCCCCCC 00000-0 00000-0 0 DDDZ 2 AAAAA EEE.EEEE FFF.FFFF GGGGGGG HHH.HHHH III.IIII JJ.JJJJJJJJKKKKKZ KEY: A-CATALOGNUM B-EPOCHTIME C-DECAY D-ELSETNUM E-INCLINATION F-RAAN G-ECCENTRICITY H-ARGPERIGEE I-MNANOM J-MNMOTION K-ORBITNUM Z-CHECKSUM
For more detail see www.amsat.org/keplerian-elements-formats/
If your trial verbose file works OK, it might be worth it to have one of the high school kids write a script to convert from TLE to verbose - it is a pretty straightforward exercise in text manipulation.
The TLE format is actually the older format. AMSAT came up with the verbose format in order to facilitate distribution on voice networks. I fondly remember copying them as read by Rip, WA2LQQ on the weekly Tuesday night75 meter net. At the time I was not really all that fond of it since a static crash at the wrong time could wipe out a critical digit and render an entire element set useless. But it was the only practical means I had to obtain them as a broke high school student in the 1980's, at least until I discovered that the government would mail them to me once a week if I asked nicely :-)
de KM1P Joe