Opening and closure of certain mechanosensitive ion channels have recently been linked with the presence of lipids in or near their pores. Although non-conducting structures of mechanosensitive Piezo channels do not show the presence of lipids in the pore, computational simulations suggest whole phospholipids enter the Piezo1 pore in the closed state. Here, to probe this phenomenon, we conduct coarse-grained (CG) and all-atom (AA) simulations of Piezo1 with different solvation algorithms and equilibrium protocols, including CG-to-AA reverse mapping from Martini CG force field to CHARMM AA force field. Our results show that the lack of initial hydration of the upper pore region, enabled by common CG but not AA solvation algorithms, allows entry of whole lipids through gaps between pore helices during subsequent equilibrium simulations. Absolute binding free energy calculations show that these lipids are thermodynamically unfavorable, indicating they are likely kinetically trapped in the pore during microsecond-long AA simulations. An alternative equilibrium protocol is proposed to avoid such simulation artifact for channels whose pores are walled with transmembrane gaps. This work underscores the notion that, as simulated systems become increasingly complex, interpretation of simulated data in physiological contexts requires extra precautions. When no experimental data is available, free energy approaches such as those implemented here appear as trustworthy validations of results observed from MD trajectories.