He's just pointing out that you seemed to be making quite a leap saying "it indicates nothing" based on a sample size of two. Based on the rest of your comments it seems like you weren't actually saying that it's a completely useless way to measure rates of suicidal ideation, but just pointing out the possibility of there being some limitations to the method and wondering out loud what the parameters were. Which I think is perfectly valid and worth looking more closely at.