I'm a developer. I looked over the Chainlink white paper and documentation, and the blog post by Google. Here's a few reasons I would not use Chainlink for my own apps:
The current "mainnet" is not decentralized. there's no way to decentrally assess if an oracle is any good or not.
It took them 2 years to basically build centralized oracles, which is extremely unimpressive (and already exists). I do not have much hope for the future or the competency of their engineers.
The white paper doesn't coherently describe how they plan on making their centralized design decentralized one day. They broadly go over their "decentralized" reputation strategies as if creating sybil resistant decentralized reputation is something of an afterthought, when it's actually an unsolved computer science problem. They've had 2-3 years and 30mil in funding to actually publish technical specifications for this and instead they decided to build trivial centralized oracles. Extremely worrying. It kind of reminds me of IOTA "we'll remove the coordinator later". No you won't. You don't know how.
News outlet and journalists keep reporting on "partnerships" with Google and Swift. But if you actually read the source for the "partnership", it's just a blog post by Google Cloud that promotes their own service (BigQuery) and shows an example about how Chainlink users can use Google Cloud if they want. In no way is Google partnered or planning on using Chainlink themselves. The Chainlink codebase is extremely trivial to rebuild if Google wanted to get into the blockchain oracle space.
While there's certainly profits to be made by trading Chainlink, I recommend you do not HODL it with the hope that one day developers will adopt it as their source for decentralized oracles. That certainly won't happen.