We show that the new data on mean multiplicities measured in p+p and A+A collisions together with the updated list of resonances lead to the significant changes of the obtained freeze-out lines. The new A+A line gives much smaller temperatures at high collision energies and agrees with the values obtained at the LHC. The newly obtained p+p line is much closer to the A+A line than previously expected, and even touches it in the region where the $K^+/\pi^+$ horn appears in the data. It indicates that the temperatures that will be obtained in the beam energy and system size scan by the NA61/SHINE Collaboration might be very close. However, our analysis shows that the chemical potentials could be very different for the same energies in A+A and p+p. It adds more puzzles to the set of surprising coincidences at the energies close to the possible onset of deconfinement.