The tough part is figuring out what the MVP actually is for your situation. It might be a Shopify site that just sends you an email when someone bought your widget, or it might be a multi-year R&D effort that ends with Falcon 1 being launched.
I also think that the "Lean Startup" concept is often abused by people to justify cutting corners on things they should not. The prime example is paying zero attention to code quality and testing prior to launch, then failing to get traction because the product is so buggy and unusable that the customers don't stick around. Don't over-architect and over-test something that may not sell, but some amount of design and proper engineering is necessary to ensure that you have something worth buying.