Ignoring the huge cost of R&D (the sunk cost you refer to) is disingenuous. It isn't even free to produce the copy of the software. There is marketing, bandwidth, upkeep. Maintaing backend servers and so on.
I agree, for the most part the price of the software doesn't have a lot to do with the cost to develope it, but you cant ignore the R&D costs.
I’m not ignoring the R&D costs. But a sunk cost is a sunk cost¹; wanting to charge more for software because you worked hard on it is a sweat-of-the-brow argument, and it doesn’t work. Nobody cares how much money was spent on the R&D which went into some software. People care about what they can use it for.
I mostly agree with you, but to the extent that your software competes with other, similar software built by companies with similar cost structures and addressable market, the cost of that sweat factors into their costs as well, so does have a (loose) connection to the viable costs for that software in the market.
If by “software” you mean off-the-shelf proprietary products, the price you pay for it is entirely an arbitrary construct chosen by checking what other market players charge for similar software and seeing what buyers are prepared to pay. Absolutely no other considerations go into pricing a software product, since the software is literally free to produce, plus the cost of packaging.
(I could, for completeness, as a clarification to the latter point, mention that the cost of hiring programmers etc. does enter into the equation at the start of a proprietary software project, but once the cost of development has been paid, it’s a sunk cost and no longer relevant for making any pricing decisions.)
See also:
https://www.joelonsoftware.com/2004/12/15/camels-and-rubber-...