The first document explains that people who take risks (like riding at high speeds or on a mountain bike) DO wear helmets in the Netherlands. Which is the reason why helmet wearers show up in hospital more often.

The second is a recent paper that applies evenly to people who ride a bike "normally", [From experience: at low speeds on segregated infrastructure at around 15 km/h or so]. My interpretation is that it concludes that wearing a helmet would improve safety for cyclists somewhat; however it would not (currently) be risk-cost-effective according to their measure; and would require intervention to break even.