"Public goods" like parks, museums, bridges, roadways, transit, nature preserves, community spaces, and public safety services produce both direct value to their immediate users as well as substantial diffuse value to their community. Direct value can be captured by user fees, tolls, subscriptions, etc but capturing diffuse value is challenging. A park raises surrounding property values even for people who do not visit the park. Good transportation infrastructure increases the value of surrounding land and and productivity per capita even for nonusers. Relying solely on user fees may force some of these entities to close or fall into disrepair, thereby reducing overall value by substantially more than it would have cost to maintain them. And in some cases shifting the cost burden to direct users substantially lowers the diffuse value, for example back when fire fighting companies would let houses burn unless their owners paid them, ultimately resulting in more overall community fire damage.
In these cases, subsidizing these public services with taxes (optimally Georgist land-value taxes) is an economically rational decision.
One could plausibly argue that artists similarly produce diffuse value e.g., raising the profile of their nation or culture, making their neighborhood a more desirable place for people with money. Not only do artists typically struggle to collect a share of this diffuse value, as renters the very value they create often ends up pricing out of their community. I could imagine cases where it is a net benefit for a government to subsidize such entities if such subsidy is less than the fraction of the diffuse benefit that ends up being collected by taxes.
I have no insight as to whether this scheme in particular is net positive, please see sibling posts for that. I'm just explaining that such arrangements are both economically rational and extremely common in high-functioning societies.
