Depends a little bit on the purpose of the description, the tissue, and whether I'm talking to another pathologist/attending around the scope or I'm writing a report.
First, I definitely agree that having a system based on the organ/tissue type you're looking at is very, very useful. Usually, one system for "how to look at" the tissue, and another for "how to describe" -- but generally based on how I look at it, so the two heavily overlap. Simply that you don't always describe everything you're looking at/looking for.
Second, you can usually have just one general system for tumors (once you think you know what it is, you can think about looking for things unique to that tumor). Epithelioid/glandular/mesenchymal etc. lesion with a _ architecture, non/invasive/poorly defined etc. margins, patchy necrosis, whatever until you talk about nuclear changes and mitoses. I was never a fan of lengthy half-page plus microscopic descriptions on formal reports, either -- relevant positives, and only negatives when there's a strong pertinent reason (such as tissue submitted for possible evidence of infection, comment on the absence of inflammation).