It's extremely hard to know ahead of time which products will ultimately be reliable and which will be flaky. Most flaky components weren't designed with that in mind -- there are some exceptions, but maintenance headaches are normally a mistake by the manufacturer. And even setting that aside, a big part of "cheap to maintain" is how many units you ship: part of why it's cheap to maintain a Civic or a Corolla is because they made millions of them, parts are everywhere, and every shop knows how to deal with them. But knowing up front if your product is going to be a hit or a flop is just not generally possible, even though that has a huge impact on maintenance costs.
Even in a world of angels, I don't know that this idea would work in general. Maybe if it was specific to some industries where the costs are easier to estimate?