Thank you for providing enough information for us to reproduce your case.
Your file is almost like a NMEA file, only Every Single Line has an incorrect checksum. Whatever created this file went to all the effort to include a checksum (it's like a magic square: I say "there are numbers one two five and the sum is eight" you know you got all the data correctly and that a burst of noise on the phone call didn't make you miss a number or add a number. This mattered a lot in the serial days and matters now on things like data loggers and messed up SD cards.
The thing is, if someone goes to the effort to put in a checksum, we're going to use it and trust it. So we throw away every line of your output. GPSBabel, however, honors the case where there is NO checksum. If we edit your file to remove the *XX (where XX is a two digit number) from the end of every line (If you're a vi user, that's :%s/\*..$//, but the basic trick can be done with about any text editor.) we read that file just fine.
You really should contact whomever makes a Solmeta and tell them to read https://en.wikipedia.org/wiki/NMEA_0183 and search for the word "checksum" - they even include example code to do it right and checksum checkers and calculators are plentify. (Hint: the "83" part of that is "1983" - this isn't exactly hard to get right.
Interestingly, I see that "According to the official specification, the checksum is optional for most data sentences, but is compulsory for RMA, RMB, and RMC (among others)." - in the spirit of Postel's Law of being liberal with what we receive, we actually treat it optional everywhere - but if you have it, it has to be _right_.
As a side note, what do you have that uses Delorme GPL? That's a format that I thought we could drop years ago, but haven't just because that wheel hasn't squeaked.