The "stylized facts" are just data-based statistics used to "calibrate" modern macroeconomic models as well as to evaluate them.The "Real Business Cycles" approach to macroeconomics took off during the 90s, in particular when it lead to "the New Keynesian General Equilibrium" approach. Both approaches were not prepared to estimate the "deep parameters" behind all those agents' optimization problems (consumers, firms, investors, governments, central banks), so they proposed to "calibrate" their models using the statistics that can be computed from aggregated data (nowadays, to calibrate the new heterogeneous-agent general equilibrium models, we should also add those computed from microeconomic data). All those data-based statistics, like standard deviations, correlations, means, etc., are jointly known as the "stylized facts" any model should replicate. There are even estimation procedures based on these calibration ideas.
Empirical observation used initially to construct a theory. Stylized facts are empirical regularities in search of theoretical, causal explanations. Stylized facts are both positive claims (about what is in the world) and normative claims (about what merits scholarly attention). As used by Kaldor to deal with the problem of generating macroeconomic theory in a world of imprecise macroeconomic observation.Hope this helps, but there are a lot of literature that you can find in this.