Earlier quoted context omitted.
But the point is to offer rides for free and use other money (presumably tax money) to maintain (and hopefully invest in more) infrastructure. I don't see how the current quality of the KC infrastructure is relevant: if it's bad, then they clearly the $9m in fare revenue isn't cutting it, and they already need tax revenue to make it better. If they can't or don't get the tax revenue they need, then that's definitely…
Or, you could use the tax money that you could use for free rides (that only makes bad service marginally more useful) and put it towards more useful service? It's not as if this money came out of a box labeled "only for free fares." Right now this is a $9M opportunity cost, annually. To get both the free fares and $9M of better transit would double the cost of this, but then you'd have to ask why not use all $18M on…
But then there is a reason for that -- spending has diminishing returns. If you have some money and you could use it to serve an area with 1500 riders then that may be worth it, but once you've served it, the next best unserved area which costs just as much to serve would only have 750 riders, which may not be worth it. The point where you stop is where the value of the service no longer exceeds the cost.
And the value of eliminating fares has a major comparative advantage because it allows you to eliminate the cost of collecting fares, which is typically a huge fraction of the money collected in fares. Especially when the fares are already subsidized (which they are most everywhere), because then you still have the pay the full cost of the collection infrastructure but generate much less revenue. In some places the cost of collection can exceed the total revenue from the fares. And even if you beat break-even, the little amount left over is hardly worth losing the benefits of having fewer cars on the road.