While I understand the desire to get the big picture, your method does have a big disadvantage in that each card does not force you to recall every single detail. When you get a card about disease X, then you think to yourself 3 key facts about disease X very readily and 1 after a little more effort, and then you flip the card and it has 5 key facts about disease X and 2 less key facts, do you select 'again' 'good' or 'easy'? Any selection creates an inefficiency because it will require recall of the facts you do remember more often than is necessary or it will allow you to forget that 5th fact more easily. You also spend more total time reviewing and trying to come up with the big picture every time than you would if you broke it up. Getting through cards becomes more boring because you can get stuck on one card for like a minute, as opposed to pounding through the cards rapidly.
The solution is to do the following. With the 5 key facts on disease X, you make 5 cards with more specific questions, e.g. what is the neuro symptom seen in X syndrome, what is the drug used to treat disease X. Then on the answer side, you list all five facts or include a visual mnemonic or something to that effect.
This is the way Firecracker works - it asks you a specific question and then shows the answer in the context of the original lesson.
This is the way I make cards now. I learned my lesson from a time when I made cards that required recall of 5 things at once, but it was very difficult/steep at the beginning and became highly inefficient once the cards matured because I'd forget one but not the other four facts and it necessitated relearning all five together.
There can be exceptions to the rule of splitting up lists, such as when you should remember a classic triad of signs/symptoms, but the idea is to keep a high bar for what lists you make into cards.