I don't use Paw (I'm a Linux-on-the-desktop user) and I have no idea who they are but a one-time $50 cost doesn't bother me if it's something that is a key tool for my job and is demonstrably better than its competitors in the market.
In other words, if it's good enough for me to spend $20 on a license, it's good enough for me to spend $50.
True, but does this do things better then postman? What's my migration cost? For that matter, why not just use curl?
There is of course, value, but there $50 price point I reserve for things absolutely critical, with little good competition - for Example Tower, or OmniFocus.
In other words, if it's good enough for me to spend $20 on a license, it's good enough for me to spend $50.