The textbook answer to why monopsony is unhealthy is monopsony suppresses investment in production. In this case, the risk is dissuading authors from writing by suppressing their earnings.
Imagine a simplified supply chain: authors distribute through Amazon to reach readers. Amazon exerting pricing pressure on authors benefits readers in the short term by reducing their reading costs. But by reducing authors' profits, Amazon dissuades new entry. Authors with skills elsewhere write less; the diversity and quality of new books suffers. Monopsony, when it reduces producer profits below equilibrium (e.g. what these authors would earn with many Amazons fighting for their business), dissuades investment in production.
Reality is complicated by the middleman, publishers. Now we have authors contracting with publishers to distribute through Amazon to reach readers. There are multiple publishers competing for authors, so for now the authors don't appear to be squeezed as much as the middleman. Authors' continued use of publishers hints at the value of their services, which may be nothing more than collective bargaining.
Removing legacy middlemen is healthy. Skewing producers' bargaining rights is not. Disentangling publishers' rents (the fraction of publishers' profits which come from them being, before Amazon, authors' only choice for distribution) from authors' costs of production (what good authors need to be paid to compensate them for the time, skill and risk which goes into writing) is difficult.