I think they saw how SpaceX was using Starlink as launch lever to provide SpaceX a baseline of regular launches at bare-minimum cost. As RocketLab starts to scale up, being able guarantee a minimum number of launches is a significant hedge against the dips in the global satellite market.
Also, RocketLab builds their own sats and can add the Iridium constellation replacements to their order book. It's a win-win. A smart move by Peter Beck and his team.
If you can afford to outright buy a customer to secure future demand for your product, does that not imply that this customer is "too small" to matter all that much? Because you could've just used the capital you spent on acquiring him on riding out demand fluctuations instead?