One thing that Hive needs better support for is custom_json specific queries. The only way to get a custom_json is by the transaction_id, but what if I want to access them based on a field inside the custom_json or based on the id? You could index them all yourself (i.e. https://github.com/emre/hotdog), but IMO this creates the problem of centralizing this data (if your storage gets hacked, the data is invalid)
For example, even if I host a hived and hivemind node, I still can access other nodes for the same data and in case of important data, I could even access multiple at the same time and choose the data which is consistent across the majority.
So, if we could create a public node specifically for custom_jsons querying, that people could run similarly to hived and hivemind node, that would be amazing.
Since worked on this problem sort of, maybe you two can work together on it. Would be much appreciated, and even something I would def. vote for via DHF.
Edit: This could also be useful https://github.com/inertia186/meeseeker
RE: Microservice idea