We (almost) ALWAYS explode our blocks, once we insert them into our dwgs. - I thought everyone does this, (?).
AAACK!!! May a herd of Yaks infest your lower colon.
Using blocks for repetative elements reduces files size, sometimes drastically. A steel structure using unit blocks for the members may have a file size of 300K while the same file using 3Dsolids would be 2MB.
Blocks, even without attributes, can be extracted into intelligent data. The unit blocks in the structure above can be extracted, using LISP, and the length of each beam (X scale) and height of each col (Z scale) can be multiplied by the member weight (number following the "x" in the block name) for a quick material weight.
Adding attributes to blocks allows a much greater degree of intelligence extraction using tools like EATTEXT. Simple things like furniture can carry a manufacturer's part number, location, phone number, or anything else you may need. A manager can then extract the data and find any piece of furniture he likes.
Simple blocks like section marks or detail bubbles can be quickly redefined to match client standards, and updated over an entire directory in minutes.
Exploding blocks around here will get you "knee-capped"