if someone needs Tomcat ... they should have enough experience to build their own stack...
Depends on the nature of the need. I 'needed' it to investigate a project that had a JSP aspect to it to try and understand what was going on. Sadly in the real world you often do not get full/willing cooperation from all parties involved on a project due to politics, personalities, language barriers, etc.
Now, while I've been in the business for - oh, let's say a long time, don't want to date myself - I'd never had any need whatsoever to deal with JSP in any way shape or fashion. For investigative purposes, I didn't need to know the first thing about building a Tomcat stack to get the information I needed. It turned out to be useful as-is, I got the essential information I needed to move forward. In this case that the system wasn't a standard JSP system but in fact a variant, which in turn led me to find out which variant it was, and now I can converse more intelligently with my client's server team about their system and how my work (front end, on this particular project) interfaces with theirs, while at the same time demonstrating that I am capable of getting information if they do not provide it. This may seem like a pointless meatbag (to borrow a phrase) trick but in the business world it can play a very useful role in "facilitating communication".
As much as I'd like to be able to specialize in everything, the practical reality is that I cannot, I am limited by time. The skill I provide is the ability specialize in what the customer needs, when he needs it, to the degree required, in a reasonable amount of time. To a nontechnical businessperson, there is real and often substantial cost in figuring out what knowledge is needed and then to find an obtain that talent - indeed there is an entire industry built to serve just that purpose that many here are probably familiar with.
The product (XAMPP) was able to help me fill that knowledge gap in a reasonable amount of time and was therefore a worthwhile effort. In the process though I burned an extra half-hour that didn't need to be burned, due to the configuration issue, and thought it worth a mention. Testing is an essential component of risk management and while that is not necessarily a technical priority it is almost always a business priority, and a strong regimen of testing leads to wider acceptance by the business community.