What are the error bars? Don't say 37,000 to 361. That's the variation between different theoretical scenarios. But real results have to be based on experimental inputs, not just theoretical guesswork.

The usual way of computing error bars is to look at the statistical variation of the data, to infer a distribution of likelihood for the results. Problem is, we have only ONE data point (i.e. us), so the variance of the inputs is infinite.

Even if you estimate 37,000 civilizations in our galaxy, how many are within 100 light years of us? Based just on the ONE data point we have, they would have developed radio and TV within the last 100 years, so we could only hope to detect them if they are within about 100 light years. Plus or minus infinity.