Reduce the pin concurrency on Protocol Labs' clusters

[RubenKelevra]

While table_loading_mode and value_log_loading_mode are set to 0 for devices with low memory, the concurrent pinning will cause large amount of memory usage, especially if the users are running multiple cluster follower.

Therefore, I like to propose to set the value to 1, as it can be easily overwritten by an environmental variable: ipfs-cluster/ipfs-cluster#1699 (comment)

This are the changed service files:

• Filecoin Params: /ipfs/bafk2bzacea2jt5lbnbhk72vtmqwrgn3ygnwrrt4lt7iuobjnnzqlghilhra5k
• Project Gutenberg (Spanish): /ipfs/bafk2bzaceapwmzqo46ozamyzcx2isvsxv4csbmq4htmlcayxywddyama7rgem
• IPFS Websites: /ipfs/bafk2bzacedbxtum5nnm4lluuuva2d6vh4hojgdhyalegmbhzcf5aacejjfbrc
• Wikipedia: /ipfs/bafk2bzaceapbip5fxcwix6znd4fvua33ge5vep4axncig3zmyontlniuc5xge