After almost 20 years of hunting, only about a dozen hot corinos, hot regions enriched in interstellar complex organic molecules (iCOMs), are known. Of them, many are binary systems with the two components showing drastically different molecular spectra. Two obvious questions arise. Why are hot corinos so difficult to find and why do their binary components seem chemically different? The answer to both questions could be a high dust opacity that would hide the molecular lines. To test this hypothesis, we observed methanol lines at centimeter wavelengths, where dust opacity is negligible, using the Very Large Array interferometer. We targeted the NGC 1333 IRAS 4A binary system, for which one of the two components, 4A1, has a spectrum deprived of iCOMs lines when observed at millimeter wavelengths, while the other component, 4A2, is very rich in iCOMs. We found that centimeter methanol lines are similarly bright toward 4A1 and 4A2. Their non-LTE analysis indicates gas density and temperature ($geq2times10^6$ cm$^{-3}$ and 100--190 K), methanol column density ($sim10^{19}$ cm$^{-2}$) and extent ($sim$35 au in radius) similar in 4A1 and 4A2, proving that both are hot corinos. Furthermore, the comparison with previous methanol line millimeter observations allows us to estimate the optical depth of the dust in front of 4A1 and 4A2, respectively. The obtained values explain the absence of iCOMs line emission toward 4A1 at millimeter wavelengths and indicate that the abundances toward 4A2 are underestimated by $sim$30%. Therefore, centimeter observations are crucial for the correct study of hot corinos, their census, and their molecular abundances.