Needs Assessment with the Delphi Technique in the Medical Field

I have used Delphi in my research and been invited expert in another.

From my point of view as a researcher, the most important issue is about expert selection. Since an answer of Delphi is usually prediction of something or opinion on something, inappropriate selection of experts could lead to significantly bias answer which cannot be easily corrected.
The second issue is the appropriateness of the response form and channel of communication among experts and researcher. Opinion of experts should be collected effectively, then analyzed and summarized fairly, and returned to experts wisely. The response form and channel of communication, therefore, should be user-friendly, flexible (for adding other unlisted issues or for different styles of experts' feedback e.g., prefer to talk than writing), and easily to see others' reasons from previous voting round.
The number of voting should not be limited to 2-3 rounds, but should be based on the results and pre-specified endpoint. However, mine had to stop with 2 rounds because of time constrain since each round took more than 2 months waiting for all experts' feedback!!
As an expert for another researcher, the research question(s) and/or objective(s) should be clear and understandable for selected experts. The feedback from previous voting round should be presented anonymously with proper format and style.