It's true that the temperature on the Sun will split water molecules very rapidly. However, the research you referred to discovered trace amounts of water on sunspots, not an atmosphere of water vapor (or worse: oceans)
The water molecules form then are broken down by the temperature very quickly. It is not that no water vapor can form on the Sun, rather the conditions are unfavorable thus the quantities are very small.
For reference, at 2200C, around 3% of water molecules are dissociated. At 3000C, around half of the molecules are split. From a chemistry POV, this thermal energy is required to overcome the bond energy within water.
The rate constant of the reaction is k=Ae^(-Ea/RT)$k=Ae^\frac{-Ea}{RT}$, where Ea is activation energy and T is Temperature. A is a constant based on the reaction and R is ideal gas constant. You can see the correlation between temperature and rate here, k will never reach infinity as long as the temperature is finite.