One of the long-term goals of exoplanet science is the (atmospheric) characterization of a large sample (>100) of terrestrial planets to assess their potential habitability and overall diversity. Hence, it is crucial to quantitatively evaluate and compare the scientific return of various mission concepts. Here we discuss the exoplanet yield of a space-based mid-infrared (MIR) nulling interferometer. We use Monte-Carlo simulations, based on the observed planet population statistics from the Kepler mission, to quantify the number and properties of detectable exoplanets (incl. potentially habitable planets) and we compare the results to those for a large aperture optical/NIR space telescope. We investigate how changes in the underlying technical assumptions (sensitivity and spatial resolution) impact the results and discuss scientific aspects that influence the choice for the wavelength coverage and spectral resolution. Finally, we discuss the advantages of detecting exoplanets at MIR wavelengths, summarize the current status of some key technologies, and describe what is needed in terms of further technology development to pave the road for a space-based MIR nulling interferometer for exoplanet science.