In contemporary spoken Latin, such as (I think) occurs among canon lawyers in the Vatican and at Latin-only conventicula, do people clearly lengthen the -ā at the end of first-declension nouns in the ablative case? If not, does this cause confusion?
tl;dr: as the risk of mistake is higher than for other suffixes, in contexts where analyzing the cases is difficult (like chanting psalms in a fast pace) people often distinguish the length less for -a/-ā (or -us/-ūs in 4th declension) than for other suffixes.
I am a Dominican friar; in our priory in Olomouc in Czech Republic, we pray parts of the Liturgy of the Hours in Latin. Of eight friars, the one most fluent in Latin pronounces all the lengths mostly correctly, few others (including me) try to pronounce them making many mistakes, and most don't care at all or at least in difficult cases like the -a/-ā suffix. Since our native language distinguishes long and short vowels, pronouncing everything short sound bit strange to us, but for those unskilled in Latin, the only alternative is to pronounce it "randomly".
The main factor is native language of the speaker - lengthening is very rare among those who are not accustomed to it from their native language. The factor number two is how we have been taught - even when distinguishing lengths is natural, their correct pronuntiation is often not emphasized, so we learn just few basic rules, and even though spotting the difference while listening is no problem, few people rely on it. The third factor is the typical occasion to speak, whether it's teaching or conversation or prayer. Even though I try to pronounce the lengths properly (especially nominative vs. ablative), I often err even while reading or speaking slowly, and while praying together, I focus on the melody and then general meaning and I most often lack the time to process the cases. People whose main concern is not recitation of psalms but teaching or communication in the language usually can be relied upon concerning the vowel lengths.
I don't know about the Vatican. But I've met very few people at conventicula, living-Latin events, etc., who make any distinction whatsoever. I don't generally have a problem, I think in part because nobody talks in insane periods like Cicero uses for orations, and with many speakers, unfortunately, though by no means a majority, word order is closer to English than to Latin, so I don't have to hold the possibility of nominative/ablative in my mind while I wait for a verb.
I know of two people who actually lengthen the vowels as the Romans did—that is, who pronounce long vowels over a longer period of time than short vowels. I know Finnish Latin speakers often observe this rule, since Finnish has vowel lengths. I haven't listened to Radio Bremen in a while, so I don't know if they do or not.
The thing is, though, that I'm very unused to listening for lengthened vowels, so on the rare occasions when I hear people use them, it basically makes no difference whatsoever to me, at least as far as I can tell. It's possible that when my comprehension is more secure or I've just had more experience listening that will change, but for now I don't really even notice.
In contemporary spoken Latin in Finnish all vowel quantities are carefully articulated. There is nothing special about the first declension ablative. I have therefore learned to expect it, and it will be easy to confuse me by ignoring vowel lengths in pronunciation. The Latin news broadcast Nuntii Latini is a prime example of Latin spoken in Finland, but it can be heard in other contexts as well, including certain academic events.
I had a Latin teacher who insisted that the long a at the end of ablatives of first declension nouns be pronounced for a noticeably longer time than other vowels. This was the only long vowel she insisted upon and she made a point of it, not only telling us that it was significant, but requiring a very exaggerated lengthening. I'm not qualified to have an opinion about whether this is important, but she made such a big deal of it that when I read to myself I still, 10 years later, attempt to observe the distinction in my subvocalizations.