In the last decade the challenging analysis of previtreous behavior of relaxation time (τ(T)) in ultraviscous low molecular weight liquids led to the conceptual shift of the glass transition physics toward theories not predicting a "finite-temperature" divergence. This "breakthrough" experimental finding was strengthened by the discovery that "dynamic" (i.e. from τ(T) fitting) and "thermodynamic" estimations of the "ideal glass" (Kauzmann) temperature do not match, what in fact questioned its existence. In this report, due to the novel way of analysis based on the transformation of τ(T) experimental data to the activation energy temperature index form, the clear prevalence of the "finite-temperature" divergence is proved. The obtained "dynamic" singular temperatures clearly coincide with "thermodynamic" estimations of the Kauzmann temperature, thus solving also the second mystery. The comprehensive picture was obtained due to the analysis of 55 experimental data-sets, ranging from low molecular weight liquids and polymers to liquid crystal and plastic crystals.