I work for a consulting firm and we typically provide Arc Flash Studies with our new projects that we do for clients. Currently, we are trying to approach new/existing customers to perform an Arc Flash Analysis on their facilities if they do not currently have one. Of course, with every big client, it comes down to two things: Cost & Is this required or just a benefit to us?
I know in reading OSHA & NFPA 70E the standards make it sound like a requirement, but does anyone have a specific statement that would confirm that an actual Arc Flash Study (Visual Inspection of gear, use of an arc flash study software (SKM/ETAP), production of arc flash labels) is required by OSHA/NFPA 70E?

OSHA treats it as the general duty clause.

https://www.osha.gov/news/newsreleases/region1/07092012
https://www.osha.gov/news/newsreleases/region1/10212013

Enforcement actions have been going on for quite a while for arc flash. The "shot heard round the world" was around 2006 when OSHA fined a major automotive company (I don't recall if it was GM or Ford) for arc flash citing 70E as their reference guide.

70E is not per se a requirement in the same way that other standards are. There is nothing stopping a company from wiring their own standard and indeed some have actually done that. But the big advantage is that it is a recognized consensus safety standard. As such it is a strong defensive argument to claim that a company is following a standard (if they are). Thus OSHA would have to prove that the consensus safety standard is deficient which is a much harder argument to make because they would have to reference something better. If a company "goes it alone" then they have the burden of proof that their standard is sufficient from a technical point of view which is a much harder argument to make and much riskier than following an existing standard such as 70E or NESC.

Paul pretty much-summed everything up already. I have written an arc flash that might help you with this: -
Thanks

In re-reading it I'm not sure that an arc flash hazard study in the form of IEEE 1584.1 is as strong of a case as NFPA 70E and that's really what the OP is asking about as compared to NFPA 70E. 70E clearly gives two different methods for doing an arc flash risk assessment (using the right terminology here): a table based approach and an "engineered" approach. Annex D has at times referenced up to 8 different hazard assessment models. Annex F sadly gives a very incomplete risk assessment procedure and makes no reference to other more complete ones. So in terms of the risk assessment, end users are on their own when implementing the engineered method.

So in terms of doing an arc flash hazard study as OP describes, no, it is not required by 70E at all. In fact one could just implement the table based approach and be done with it. OSHA requires SOMETHING to be done and in some ways one could state that this is a minimum code standard. There are obviously problems with the table and lots of scary footnotes. But coming from someone that travels to all kinds of customers many of which have done little to nothing in terms of arc flash studies, the table based approach may be all that particularly contractors have available in many cases. And many of my customers are very small municipalities or private businesses were the cost of an arc flash study is larger than their entire annual profit margin.

For all these reasons NO I can't point to anything that would endorse doing an IEEE 1584.1 style hazard assessment and say a LOPA style risk assessment because neither one has provisions for small customers. Until you get the engineers out of the way except when it is appropriate, it will continue to be an uphill battle trying to get standards to support any position mandating engineering studies.

Let's face it this is the reason that for instance NFPA 85 and 86 are the way that they are and the reason that they still carefully tiptoe around very small burners where a different standards body actually supplies a Code. ISA is desperate to make NFPA 85 and 86 Committees go to IEC 61511 for burner safety systems but NFPA 85/86 has steadfastly refused citing the fact that IEC 61511 contains no provisions for small applications as an example.