This is a problem I struggled with for three years. My background is computational physics, and like you I have many active projects at any one time. My work evolves three main activities : scientific research, software development and paper writing.
During my first post doc position of three years duration, I managed to come up with many different ideas for research and features for the software package I was developing, but simply lacked the time to progress with them all. I flitted from one to another, trying to devote a little bit of time to each idea, but never making real progress on most. I also never kept a proper lab book, just having a directory on my computer for each project. This meant that I often left work untouched for weeks, and couldn't remember what I did last, often having to rely on time stamps to see which data was newest - very unsatisfactory indeed.
However, I recently made two changes which have made a massive difference to my ability to effectively deal with many projects. The first is dedicating large blocks of time (3 hours generally) to a single project and focussing on that one task exclusively - no email, no meetings, no Internet if I can avoid it. The other change is setting out a master plan for each month. I prioritise the best quality ideas and time critical work such as project reports, conference abstracts and talks. Other projects are put off until I have free time or they develop into something better, at which point it becomes a priority anyway.
My tool of choice from a practical perspective is Evernote. The ability to group collections of notes and restructure/combine them as needed is crucial to dealing with many related ideas which may converge or diverge at different points in time. Also when I finish a block of work on a project I religiously add my thoughts and results to the lab book, so that next time I come back to it (which potentially coud be weeks or months hence), I know exactly where I got to. I also refer to specific directories on my computer so that I can always recreate the data/figures at a later date, without having to hunt for the right directory.
When I have a new idea I start a new notebook under Research, which contains a few bullet points about the idea - enough that I can recreate the thought should I completely forget about it (it happens!). Later I will usually come back and add a project plan, from which point things are usually pretty linear. Given each idea/project has its own directory of notes, I always have an overview of what is going on, while the linear series of notes catalogues the thought processes and data. Using this approach I find it much easier to keep control of my research and its direction, and I am much happier and productive for it.
For me an electronic lab book is essential. As others have mentioned if you travel a lot then physical notebooks are suboptimal. I still have a book for quick notes and equations, but they get typed into the lab book/LateX as soon as I get to the computer. Although my lab book is currently quite small, as it grows I want to be able to search for data and text. There is of course a potential downside with my current solution - it relies exclusively on the existence of Evernote. If the company fails for some reason then I potentially lose all my notes (since they are hosted on their servers). I am always on the lookout for new software which does the same thing, and preferably open source which takes away the heavy reliance on a single commercial entity.