On the possibly low H2 formation rate in dense clouds