Probably the bonding in $\ce{UF6}$ has significant ionic character; as with many compounds neither the extreme of "all covalent" nor "all ionic" would really be accurate.
Ionic character in $\ce{UF6}$ need not imply the high melting and boiling points we commonly associate with ionic compounds. We have to reckon with the 6:1 stoichiometric ratio of proposed anions to cations -- such an extreme deviation from equality makes discourages forming a strongly stable lattice (whether ionic or covalent with delocalized bonds). An ionic model of $\ce{UF6}$ could be an ambient gas of neutral clusters similar to sodium chloride or lithium fluoride at higher temperatures.