These are just some of the thoughts and questions I had. I'm skeptical that a landing page alone can do much since it isn't really solving a problem for the user.
But depending on the product, It could be little more than a more elaborate form of customer interview (asking someone: would you pay for this? Does this sound like something you would use?), or it could be actually an effective sales page.
Let me try to give you a couple of examples to clarify: if your product is trying to solve a moderately big and complex problem, let's say personal finance, team project management, home security, etc. With a landing page that only contains a product mockup and asking for an email, you are effectively only gauging initial interest. The thing with this kind of feedback is that an explosive, huge amount of initial interest does tell you something meaningful (You are onto something, keep going!), but moderate or low amounts of feedback don't necessary tell you much about your product-market fit or predict success. It could be that the idea sounds initially interesting but your execution is poor, it could be that your product sounds really cool but fails to solve a definite problem in a definite way, or it could be that there is a real need for your product in the real world but your landing page/marketing copy sucks, resulting in poor email engagement. In this case, the landing page should be used as one tool among many to de-risk your startup early on, along with traditional market research, customer interviews, etc.
But let's say your product is an online programming course on Flutter, or an Indian food cooking guide. You want to know if there is a market for this. If you design your landing page in a way that promises delivery of the product very soon, even if you have not yet started working on it, then your customers are already in "show me where to send the money mode" and the feedback obtained from the landing page is actually meaningful. In this case, the gap between the landing page and making a sale is much smaller.
This is just my take on this, having thought about this specific issue some time ago.
Heck, while building OnlineOrNot (https://onlineornot.com/), people weren't willing to try it out until several features beyond what I'd call an MVP were built (and it has a free tier).
IME, it's very useful to force you to clearly state your value prop. Then you can test if that prop resonates and lands with a specific audience that you generate.
It's not going to get you anywhere on its own. It's one step of the puzzle, but an important one.
If you don't have an audience already, then you have to try to squeeze getting a temp-audience into a very compressed amount of time, and you'll have relatively/very few emails.
I think part of it is just doing it - you put it up, you hustle hard for a few weeks to get 3 addresses, that makes you realize you need to build an audience before launching next time, etc.
No personal experience, just my take.
I'd rather see a survey form than a landing page. If you want to collect e-mails, be honest about it. When a page has a sign up button that goes into a "waitlist", I assume it's bullshit. It falls into the overpromise, underdeliver category and I won't buy it when it comes out.