I need to keep propagating my new work into various files.
This sounds like a job for XREFs. Blocks within files is cumbersome at best as you are finding out.
"Get more organized" with block naming, or filing (wblocks or xrefs) is not an option, for very practical reasons. I'm designing stuff. That stuff is in constant flux for quite a while, and it's not a linear process.
I'm afraid organization may be the only option. In my non-work life, I tend to not be very organized. At work, I am meticulous when it comes to file management. Maybe if you describe your work flow in more detail, we can give you suggestions. I'm sure someone has worked in a similar fashion and can give you some pointers.
If you can't make the program work with your process, you probably will need to change the process.
I'm afraid organization may be the only option as well, though it would be of a mind-numbing complexity in my case. I believe I'm pretty meticulous as well, but some kinds of meticulous put people out of business. I can't afford to be that meticulous.
The stuff I design is all about parts, assemblies, machines, and systems. I've been using blocks lately to mirror this structure, which is pretty straightforward in principle - make a block out of a part, make a block out of an assembly that's an assemblage of blocks of parts, make a block out of a machine, that's an assemblage of blocks of assemblies which are themselves assemblages of blocks of parts... It's all nesting, all the way down.
I always aim to standardize, at any and all of these levels, but there's a process involved in getting there. And that process happens when working on real projects that need to get done in certain timeframes. Parts get redesigned when I see how the machines function as a system. It's top-down and bottom-up all the time. It's a turbulent process, and it needs to be.
I didn't have these problems in the past, though. One, I didn't use blocks as much (or wblocks or xrefs). And two, I did all my work in one file, so to the extent that I did use blocks, redefining them didn't cause these problems. Both of these factors changed recently as my projects got more complex and I now have detailers working for me. Multiple files means multiple people working on a single project at once, and blocks organize my designs so that detailers can more easily know what they are looking at.
One example of how the current process doesn't work for me is that the blocks I make as I'm designing have casual names, often randomly generated. When the design (say, of a machine) gets to the final detailing phase, accurate/useful blocks are made, which I can then put together as a system. This system is then placed in multiple contexts. I try to make sure blocks are renamed at a certain point, with a nice, formal name, once I get to an appropriately stable stage in the process. But there's just too much nesting - some nested block is going to get inserted into a different file, it will be redefined back to an old definition, and I'm none the wiser. There is no way for me to know that the information is now wrong.
