I'm an experienced chant singer, so I do have some first-hand knowledge about what might be going on here.
1) The clarity of words in chant is important. But the problem for listeners is that the resonance of the spaces in which chant is usually sung/recorded is so great as to obscure even well-enunciated consonants, especially when you have multiple people singing.
2) Different linguistic cultures developed different pronunciations of ecclesiastical Latin, so you may be running into otherwise familiar words that are just being pronounced differently, e.g., "regem" in Spanish ecclesiastical pronunciation has a hard "g" as in English "get", whereas in Italian pronunciation it's a "j" sound like the English word "gem".
3) Furthermore, even if the Latin pronunciation is the same, you may also be dealing with a foreign substrate accent, e.g., choirs singing Latin with a modern French accent. We all know how difficult it can sometimes be to understand people who are speaking out native tongue correctly but in a foreign accent, so understanding a sung foreign language can similarly pose problems.
4) Finally, the act of deliberately singing words can alter the pronunciation significantly from what we are used to in speaking. For example, consider the English word "for" -- as an independently pronounced word it often has a long O sound like "fawr", but in common speech we usually hear "fer". The same thing might be going on with how you hear the chant versus how you are used to reading/saying Latin.
As a solution, I would recommend trying to learn the texts of the chants you are listening to, and then trying to pick out the words without referring to the written source. Maybe try that a few times, and then do a few trials without looking at the text beforehand to see what you can figure out.
I hope that helps!
