Perhaps we could more easily start a list of questions by reading through a good list of keywords associated with each sign to find the contexts (areas) in which they differ. We could do the same for each planet. We could collect data about facial features, reactions to given situations, preferences for noise environment, and any number of other things. By identifying a list of contexts of variation like this, we can write a list of questions that will allow us to confirm or refute our existing ideas about the signs and planets.
Getting people to fill out the questionnaire would be the difficult part, but that is just logistics. If we have a viable tool to use (questionnaire), then I'm sure we will find some opportunity to implement it eventually. I have a website that could be used to host it and I could write the script to accumulate the information. The chartered purpose of my site is astrological research and development. I don't have the traffic to gather a lot of individuals, though, so they would have to be sent from somewhere else. Anyway, the first step clearly seems to be making a solid questionnaire.
We also must consider that a few individuals might want to intentionally skew the data just because they have too much time on their hands. So, in gathering individuals to provide data, we'll need to target a certain demographic and not leave lots of links to the questionnaire strewn about everywhere. If we create a good questionnaire and get everything ready for data collection, I could drop some emails to admins of sites like astro.com and a number of others to see if they would shoot me a link for a week or so in order to further astrological research. I wouldn't count on astro.com for cooperation, but there are many other such sites with large amounts of traffic from a narrow demographic that may be motivated to contribute.
If implemented in this way, the questionnaire would need to be more generalised. I would love to have a huge database of twin data as well, so we could write another questionnaire just for twins and then I could drop some emails to sites about twins to see if any of them would want to contribute to astrological research on themselves. All of this is doable. I would recommend starting with the general questionnaire, derived from expected contexts of variation between the signs and planets. If we do this, we will actually have data to say something about whether or not certain astrological assumptions are functionally correct.