Why does it have such a low solubility? I would have thought the OH group would make it more soluble.
Interesting question but there is a logical fallacy. You are connecting the concept of solute's polarity with its water solubility. Note the word polarity is a catch-all term for many interactions. Solubility is the hardest thing to predict or model! It is very hard to predict melting point as well as boiling point. Not sure why general chemistry textbooks only misguide students and they have to unlearn so many things at a later stage.
You made very good observation. Let us look at your list from another angle i.e., chromatography- a special type of chromatography. Lets us say we have a polar stationary phase like silica and use a mobile phase which contains acetonitrile + water at a neutral pH. In this chromatography mode, the most polar compound elutes in the last and the least polar first. If you inject a mixture of phenol, resorcinol and phloroglucinol into a column, what is the elution order?
Phenol, then resorcinol and then phloroglucinol. Lo and behold, this is exactly like your prediction! So there is indeed a trend. However we cannot correlate this single observation with the water solubility.
Let us think of a cyclic sugar, cyclofructan: 21 hydroxyl groups, yet it is insoluble in water! It is insoluble in many organic solvents as well. So it is very hard to correlate the number of individual functional groups and predict water solubility. Exceptions like these will pop up again and again.
However there are softwares, such as Advanced Chemistry Development (ACD) which can predict solubility. No idea what models they use. You can look up their documentation or even contact them.