Frankly, get going with your adwords test. There is no other dataset nor platform so comprehensively used that it's performance approximates a real market. With search data (volume, trends, CPCs, and your performance) you can extrapolate everything else from the size of the market and cost to acquire therein, to the relative CAC from other channels.
By that token, it's also the only way to get a valid sample of what comparable companies' CAC might be. Why? Keep in mind the unique context of search marketing - it is an intent based audience. That means, it's the only channel through which you can draw some valid assumptions that YOUR performance from a sales based ad on a specific keyword pointing to a similar landing page approximating that of what your comparables experience. To do the same on Facebook, email, or other channels removes the context of intent and creates any other number of use cases that result in THEIR experience advertising there differing from your experience (i.e. targeting, audience, cpc, etc.).
It's not easy to do as you further need to draw some assumptions about brand strength, existing awareness (or lack thereof), and accessibility on various platforms (you could run the exact same ad, landing page, and have equivalent brand strength as a comparable but if they are on iOS and Android and you're only on Android...).