From performance and scalability, do keep in mind that table & blob storage retrieval is conducted via API calls.
If you're pulling BIG DATA every minute from 1000's of blobs/tables, you might find that performance will be impacted. You'll need to architect a solution that uses dedicated machines for pulling the data, etc.
Networking speeds will come into consideration at scale, as will the TCP port limit of 2^16-1 or 65535.
As for which is better, I've experienced the similar results with both and found that some Azure services may require one or the other.
Blob is good for unstructured data such as photos, documents, videos, backups, and varying format logs, etc. Blob is typically recommended for "big data".
Table storage is more "SQL-Like" and is typically better for structured data, and can also be more performant if developed corrected.