To add to the SS vs CV debate, I'd like to note that in the vast majority of cases, SSs did not engage capital ships *during* naval battles, but before or after (IIRC, Shokaku and Taiho at Philippine Sea were the sole exceptions? Doing this from memory here, so I may be misquoting history.). The chief factor being the speeds of the TFs when they are engaged. CVs *do* have a window of vulnerability when they conduct air operations and turn into the wind to launch/retrieve a/c, which the US subs took advantage of (again, IIRC).
I can see the argument turning either way (perhaps with all sides being correct within their own context), but within the greater context of naval operations, ships engaged in battle, statistically, were less likely to be successfully engaged by submarines because the sub had to be in position and undetected when the torpedo launch happened (no "stalking").
There were other factors as well, such as the Japanese ASW detection equipment being less effective at battle speeds (as was UK's ASDIC before c. 1943) that made it easier for the sub to be undetected when it WAS in position ahead of a CV group that was conducting air operations. (A concatenation of ideal circumstances, i.e., luck.)
From what I read from Podcat, it sounds like PDS is on the right track by making SS attacks against capital ships somewhat random, and highly dependent upon luck, which, in the heat of battle, seems to have been a deciding factor in scenarios where SSs were able to launch successful torpedo attacks against capital ships.