It's certainly effective. I once spoke with a guy who founded a relatively large e-commerce biz ($100m+ revenue) and had been exploring a new product. To validate it, he put together an entire webstore complete with an actual checkout - not just a price, but people could actually put in a credit card and pay. He bought a variety of ads for it, collected orders, and then just cancelled all the orders. This essentially allowed him to measure all the numbers he needed to validate whether it would be profitable (ad targeting, conversion rates, etc.) without the cost of having to actually manufacture the product.

The best part of this approach is that you measure actual conversions, and your numbers aren't distorted by people who are only willing to subscribe to a mailing list, or even people who make it all the way to a checkout page before changing their minds. It's purely people who would've actually paid for your product, because they really thought they did.

On the surface it seems sketchy and possibly unethical, but I suppose the inconvenience to the customer is relatively small. They experience a brief disappointment when they see the email telling them the order is cancelled, and that's likely the full extent of your negative impact on them. You could even argue it'd be unethical to yourself / your investors to not validate an idea every way you can before spending lots of time and energy on it.