Thanks to @AtlasTSO for gathering the evidence and @ivy_oracle for your reply.
My opinion is that this is not a case of collusion with another party but rather Ivy Oracle running a second provider.
There’s no doubt in my mind that Ivy Oracle has developed their own code and can demonstrate that they run it on their own infrastructure. Reviewing that is a pointless exercise unless the ‘anon’ provider gives the same access for comparison purposes, which is unlikely to happen. I’d love to be proven wrong.
I struggle with idea there’s another random provider ‘making the same mistake’ and ‘tackling the problem’ in the same way. It’s not an upstanding provider we’re referring to, it’s an anon provider with similar success to Ivy.
‘No common downtime’ is somewhat irrelevant. If one provider can run with minimal downtime, it stands to reason their second provider would operate similarly.
The collusion tool on Flare isn’t currently useful due to the pro providers running the same sample provider code, which heavily skews the ties towards those providers. If those providers were removed, many of the relationships would be different, and the tool would be more useful, similar to its effectiveness on Songbird.
Hopefully other providers can review the above thoroughly and give their point of view.