I think you should try the Generate Near Table (Analysis) tool with its closest="ALL" option so that:
Multiple near features will be written to the output table
Otherwise, by default, only rows for the nearest features are written to the output table, as you have observed.