Facebook Insights API can give you hundreds of metrics. If you want to run a simulation to know how many conversions you can expect if you spend 2x, how do you know what are the variables that influence the conversions the most?
The answer to your predicament is: covariance. If you are familiar with probability, you know that covariance will tell you how strong is the correlation between two random variables.
In the context of ads performance analysis, those variables are fields or dimensions (age, location, audience size, and so on) and metrics (clicks, impressions, conversions.)
In the following matrix, you can see a correlation between Spend, Clicks and Reach (people reached), with a range between -1 (negative correlation) and +1 (positive correlation.)
This matrix will be different for each campaign, or however you decide to slice your data. In this example, you can see there is a positive correlation between all the metrics (the more you spend, the more people you reach, the more clicks you get.)
However, the correlation between reach and clicks is strong, which will suggest there is a high engagement between the users and your ads.