You have 10 feeds. You have 2 devices. You hit refresh on one device. You read some articles. You go to your other device. You hit refresh. You see the same articles you’ve already read marked as unread, because this device doesn’t know what you read elsewhere.

Aggregation may be useful to keep articles for the amount of time YOU decide. Blogs come and go and with it the articles. Or they move to different URLs and then are lost until you find them again.

The main reasons I aggregate are 1. Speed: feeds are refreshed on the server, so each client pulls down the aggregated feed instantly. 2. Sync: read/unread items are synchronised between clients/devices. Honestly, that's enough. But there are many other smaller gains, such as being able to filter articles (Mark as Read) based on keywords I'm not interested in.

Same reason why Dropbox isn't a replacement ftp server.