Best guess: that's because no organizations are writing their first PL/I apps in 2025; and there aren't many PL/I-using organizations who weren't big IBM customers, who largely migrated from OS/2 to Linux. So I wouldn't be shocked if a Windows version wouldn't bring in enough money to make even a relatively easy port worth it.
Next-to-best guess: author was an OS/2-head, many of whom have been hating M$ Windoze for 30 years at this point! Ah, those were good days.
Next-to-best guess: author was an OS/2-head, many of whom have been hating M$ Windoze for 30 years at this point! Ah, those were good days.