The reason maps can’t be used at a property level (to identify individual properties at risk) is that they are not designed for such purpose. None of the modelling methodology applied, in the available models, is robust enough to provide with sufficient confidence in outputs being suitable to identify individual properties at surface water flood risk.
Based on modelling methodology, data availability and parameters used in a particular location, a confidence score and related suitability rating is established (for a given model). In the national model, confidence score varies between 1 and 3 out of 5. 3/5 means outputs are suitable for identifying likelihood of flooding at street scale (i.e. which streets). National model outputs cover 98% of the country by area.
The most sophisticated, locally supplied models, achieve a score of 4/5, which translates into being suitable for identifying which parts of a street are at risk (but not individual properties).