A landing page doesn't work for everything right? A web-based product, sure, but what if it's something tangible? You need to build a prototype.
Look at a technology that has had a massive impact on the world: fracking. People pursued that for many years with limited results before they were able to do it successfully/economically.
And now we'd like them to stop, please. Fracking is a terrible idea.
If you take the YC credo, "Build something people want", there are two assumptions that need to be tested. 1) Do people want it and 2) Can you build it? I think that most times it's best to test #1 first.
A landing page will prove people will buy the product.
A prototype proves the product can be built.