In relay attacks, a man-in-the-middle adversary impersonates a legitimate party and makes it this party appear to be of an authenticator, when in fact they are not. In order to counteract relay attacks, distance-bounding protocols provide a means for a verifier (e.g., an payment terminal) to estimate his relative distance to a prover (e.g., a bankcard). We propose F lexiDB, a new cryptographic model for distance bounding, parameterised by different types of fine-grained corruptions. F lexiDB allows to consider classical cases but also new, generalised corruption settings. In these settings, we exhibit new attack strategies on existing protocols. Finally, we propose a proof-of-concept mechanisation of F lexiDB in the interactive cryptographic prover EasyCrypt. We use this to exhibit a flavour of man-in-the-middle security on a variant of MasterCard's contactless-payment protocol.