The use of pellet injection for disruption mitigation purposes has created a need to extend pellet modeling into new areas of the parameter space. This comes with a number of challenges, for instance the increased impact of effects such as plasmoid drifts and rocket acceleration, which must be understood and incorporated to obtain an accurate predictive model. I left the field at a time where several exciting research activities were conducted in this regard, with both several ideas and open questions I never had time to pursue on my own. I therefore make this force spirit appearance to summarise these ideas and questions as a source of inspiration for future activities in the community.