The point is you wouldn't endanger the lives of teenagers to build a community theater because you believe the lives of teenagers are more valuable than a community theater.

But repairing a community theater could easily cost $100,000 which would save the lives of ~22 people in Guinea if donated to certain bed net charities. And even the most die hard theater fans probably wouldn't save a theater if to do so it would require risking 22 lives for.

So in this case quantification isn't failing, it's succeeding. It's correctly telling us which donations are more effective. And all our intuition agrees that the $100,000 theater is worth less than the 22 lives.