Yep and another notable but subtle cost is the manufacturer must now keep two separate supply lines open: one to refurbish previously used boosters, and one to build new ones and components. Keeping both lines open is expensive and running either one at reduced capacity (ie. if few boosters are produced because many get reused) is a huge expense.
This leads to one of the most crucial rules in reuse: it can only work at a consistently high flight rate (they're doing so with Starlinks and Musk claims they're at breakeven) because that is what is needed to overcome the fixed costs of running two supply lines such that they it doesn't dominate expenses. Otherwise, more efficient production without reuse makes more sense.
The costs of repairs tend to be dominated not by routine refurbishment but by unexpected, severe damage sustained during flights - something which becomes increasingly common upon multiple reflights.
Similarly, despite some expensive upgrades to support reuse, it is all but guaranteed that successive reuse attempts by even F9 B5 will quickly become prohibitively expensive due to such significant damage after multiple uses.
It remains to be seen if they can keep undercutting the market.