Forecasts and size estimates shown with multiple decimal points are scary because they invariably imply a phony level of precision. When you’re estimating something, understand up front how precise the answer has to be, and present the result accordingly.
Doing a near-term estimate for a production forecast is one thing. But if the question relates to a market’s size to gauge relative market share or reasonable long-term growth expectations, it’s probably appropriate for your answer to be a range, and maybe a pretty wide one (2x or 3x differences between the low and high end may even be reasonable).
Also, rather than investing all your efforts in one estimate, approach it with multiple methodologies or sets of inputs to create credible boundaries for your estimated range. That’s “precision” that’s more valuable than any level of phony decimal places.