Here is a fun experiment you can do to show how they try to deceive you into thinking you need these additional products. It works. Just try... I think you will be a little concerned when you do.
pick a page.
Run your Google PageSpeed score on that page from the Google Dev site. Then put your Google PageSpeed API key into the appropriate place on the w3tc dashboard and have it run the same test through Google. You may find, as I did, that the PageSpeed score Google gave you when you did it on their site was 92. But when you run the same test through the w3tc dashboard, it is now 68. Wow.. Big difference considering it is the same api. Could it be that the w3tc manipulates the results? Subscribe to one of their partners and run it again (but WITHOUT configuring that component in a way that makes a difference. For example, subscribe to MaxCDN, but don't put any content there yet. Did your Google PageSpeed score as presented by w3tc magically get better? Closer to your actual PageSpeed site? Mine does. And that's probably more than just "Wong". I'm pretty sure Google api licensing doesn't allow you to manipulate the result data to falsely give the impression that your product is doing good things. What a total mess.