I think there are good odds that for any platform where the code can actually be built that function just read randomness from the OS, but you're right that it isn't guaranteed.
the bad odds are that people are good at re-inventing bad ideas by dint of hubristic assumption that they're being imaginative/original. you can bet
someone will go for security<-->obscurity simply because they have some exotic or old platform that they believe somehow protects them from attacks, and use it to generate seeds and/or sign txs offline etc. I realize this is many layers of (now anachronistic) hypotheticals in the case of bx-seed though, but these things bear repeating every so often
It's also likely bad for the non-cryptographic usage in libbitcoin. An attacker can observe some of the choices made by the software you can recover the RNG state then use it to predict the random choices that were used and will be used, which likely results in DOS vulnerabilities.
hmmm, but wasn't the low-entropy seeding limited to the bx-seed codebase, or has a similar problem been found in the libbitcoin dev toolkit?