You are assuming, of course, that landlords stick to conservative affordability limits and turn down willing tenants who fail the test, even by a small margin.
I agree this dataset is not ideal for these sort of calculations, but the figures are supported by my anecdotal experience.