Molecularly imprinted polymers (MIPs) are synthetic receptors made by template-assisted synthesis. MIPs might be ideal receptors for sensing devices, given the possibility to custom-design selectivity and affinity toward a targeted analyte and their robustness and ability to withstand harsh conditions. However, the synthesis of MIP is an inherently random process that produces a statistical distribution of binding sites, characterized by a variety of affinities. This is verified both for bulk MIP materials and for MIP's thin layers. In the present work, we aimed at assessing the effects of inhomogeneous versus homogeneous imprinted binding sites on electrochemical sensing measurements, and the possible implications on the sensor's performance. In the example of an Electrochemical Impedance Spectroscopy (EIS) sensor for the 17β-estradiol (E2) hormone, the scenario of inhomogeneous binding sites was studied by modifying electrodes with an E2-MIP polyaniline (PANI) thin layer, called the "Imprinted PANI layer". In contrast, the condition of discrete and uniform binding sites was epitomized by electrodes modified with a thin PANI layer purposedly doped with E2-MIP nanoparticles (nanoMIPs), which were referred to as "nanoMIP-doped PANI". The behaviors of the two EIS sensors were compared. Interestingly, the sensitivity of the nanoMIP-doped PANI was almost twice with respect to that of the imprinted PANI layer, strongly suggesting that the homogeneity of the binding sites has a fundamental role in the sensor's development. The nanoMIP-doped PANI sensor, which showed a response for E2 in the range 36.7 pM-36.7 nM and had a limit of detection of 2.86 pg/mL, was used to determine E2 in wastewater.