bond - How can we say whether or not an organic compound is soluble?

In the exam recently conducted by my school, they asked whether 2-hydroxyacetophenone is water soluble or not.

I thought that since there is intramolecular H-bonding in this compund, it won't be soluble. But my teacher marked it wrong. Also, when I checked it on internet, I saw in some websites it was mentioned to be water-soluble. My question is how can we say whether or not an organic compound is soluble?

Answer

The traditional rule of thumb has always been that 1 oxygen can dissolve 4 carbons. In this case, the compound is $\ce{C8H8O2}$ which makes an oxygen:carbon ratio of 1:4, so it should be soluble in water.

The source of this rule of thumb is Dr. Pedro J. Victory, SJ, who taught us Organic Chemistry. That was a long time back but I recall this rule of him. The context was liquid-liquid extraction.

I do not remember if he added a corollary for the case of esters, where you should count only one of the oxygens of every ester group. We perform extractions with ethyl acetate all the time and there the ratio is 1:2. Very probably, he did add that.

To support this rule I have searched for some solubilities at PubChem. On basis of this data, the threshold for soluble (as asked in a comment by @Zhe) would be around 10% (100 mg/mL).

                    mg/mL T (C)
acetone             completely soluble 

methyl ethyl ketone 223    25
3-pentanone          43    25
2-pentanone          46    25
isopropanol         completely soluble
1-butanol            63    25
2-butanol           181    25
1-pentanol           22    25
2-pentanol           45    25
3-pentanol           55    25
methyl ethyl ether  230    25

diethyl ether        60    25
methylacetate       243    25
ethylacetate         80    25
n-propylacetate      19    20