Good questions and observations.
I guess first, the goal of research is to take a representative sample of people and apply what you find to the larger population. So, if I want to understand coping in people with Crohn's, I'd survey 200 people with Crohn's and hope that I could say "it's likely that people with Crohn's..." I can never survey every single person with Crohn's, so I want to find a group that's diverse enough that they could in theory represent everyone with Crohn's. Easier said than done, but that's what we try to do.
Why are the questions identical across studies? Research is generally concerned about reliability (are the findings repeatable) and validity (is it actually measuring what it says it is). To ensure that this is happening, people use "reliable and valid" questionnaires. These are surveys that have been tested and re-tested for reliability and validity across hundreds (more likely thousands) of people. If it's a questionnaire that's been widely used for decades, then it's considered the gold standard to get at a certain area of study. So you see a lot of overlap. It's less reliable and valid to create your own set of questions to ask, unless you first run statistics to ensure they're measuring what you really want them to measure. While it seems like you could just look at the questions and say well yeah those are questions about depression or coping, it's not that simple unfortunately.
The 2nd point is one of the main debates within the research community, especially for psychologists. People are so complicated, how can we reduce them to check boxes? And the real answer is you can't get the whole picture, ever, in these studies. For one, it would be way too complicated. So, researchers are trying to look at just certain areas of interest. They may ask questions about related things as well (which gets at validity). When you participate in a study using questionnaires, it's a "quantitative" study. Which basically means the goal is to get numbers (or quantify) people's experiences so that the researcher can run statistics to look for relationships between variables. These can get quite crazy in their complexity. At the end of the day though, there are always other factors (called confounds) that make it so a researcher can almost never conclude that A causes B. We can make observations, make some guesses, but rarely can we say A causes B. Because as I said, people are just too complicated.
There's another type of research called qualitative research, which is more in depth and involves interviews, case studies, observation, etc. It's much more rich in terms of understanding the unique experience of the person. The problem is they're not very generalizable to the larger population. But, they are highly valuable in trying to understand an area that hasn't been heavily studied in the past.
I hope this makes sense :)
I've done both types of research, and I really enjoyed the qualitative stuff. It's so much more in depth. Unfortunately, it's not the norm in the research world. I plan on doing more, but for many people it's too time consuming, cost prohibitive, or they want the numbers to analyze versus the words.
I could probably write about 40 pages on this, but I hope this helps clarify. Any other questions, just ask.
"Only the meek get pinched...the bold survive."