Explore intake as the main i/o layer in GCRCatalogs

[yymao]

intake seems to be a pretty cool tool. Read the blog post here: https://www.anaconda.com/blog/intake-taking-the-pain-out-of-data-access

We should explore the functionalities and features of intake (https://github.com/intake/intake) and evaluate its potential to act as the main i/o layer in GCRCatalogs (i.e., to replace all readers in GCRCatalogs).

We should

• document the functionalities and features of intake
• discuss whether intake serves our needs to replace GCRCatalogs
• list the steps if we are to migrate from GCRCatalogs to intake