I'd be skeptical that would work.
A friend did a ton of user testing of improvements to a price-comparison site back when those were the rage. With some frequency the engineers would come up with a way to help people make a better buying decision. E.g., picking a TV is a problem, so they'd make a wizard that would ask you questions and then give you a recommendation. Problem solved, right?
Alas, no. Turns out most of these user guidance things wouldn't help, because people had no reason to trust the thing. They might go through the process, but their behavior didn't change. I'd expect to see the same effect with a kiosk. Most wouldn't engage, and those who did wouldn't weight the recommendation very highly.
A website would have an even deeper trust problem, and would add an SEO problem on top. Imagine a referral is worth $1 on average and you spend $0.50 on coming up with good answers, $0.49 on making sure you're on the first page of Google results, and take $0.01 in profit. You'll very quickly have a competitor with that spends $0 on research and $0.99 on being ahead of you on Google. Sure, their data will be garbage, but the page will be just as convincing to somebody who doesn't know anything, which is your target market.
I suspect the real outcome, as with many would-be startups, is that this is a feature, not a business. Somebody like Wirecutter or Consumer Reports could turn this into solid content that would be a nice addition to what they have already. They've already built a trust relationship with their users, and they don't have to specifically find people in the (very rare) moment of choosing a new medication.