It really depends on how you define "harm" and who is harmed. Classifying "not paid as much as some theoretical model claims a person should be paid" as "harm" without examining the theoretical model in detail is naive.
For example, let's assume we care about improving people's lives. It's probably a short walk to assume that more independent software companies (both big players and smaller startups) satisfy that goal better than only a few players, or one player.
A collusion-free labor market encourages players at the top to spend on labor until they bleed. Eventually, if nothing checks that cycle, they bleed out competing for a small pool of top talent, with the winner being the one with the deepest pockets who can afford to bleed longest. Then, the winner gobbles up the talent and resources of the bankrupting companies, and there are fewer players in the market (and the remaining players get to dictate labor prices on the grounds of being the only players).
Ironically, a soft salary cap can benefit the ecosystem in the long-term; employees aren't paid as much as they could be, hypothetically, short-term, but there's more job security because a price spiral is unlikely to disrupt their employer. And there's a more diverse set of independent companies in operation, likely leading to a more diverse set of software solutions and more improvement in people's lives than if salary spirals meant that only big players could afford to play.
It's one scenario among many possible, but it's interesting to observe how maybe not paying software engineers much as they could be in a completely laissez-faire market creates long-term benefits for the engineers, their employers, and consumers.