Sale-leaseback rent is set by the seller and that bothers me

Over-rented sale-leasebacks are the standard failure mode in this corner of net lease, and the mechanism is exactly as you describe. The seller is monetizing a building, and every extra dollar of rent they agree to pay themselves converts into roughly fourteen dollars of proceeds at a 7 cap. The tenant is on both sides of the negotiation, so the rent is a financing decision, not a market observation.

The way to make the gap visible is to split value the way an appraiser would in a hypothetical vacancy. At $11 market rent, that same building at an 8 cap for generic flex space pencils to about $138 a foot against the roughly $264 a foot you're paying, call it half. So you're buying real estate plus an unsecured credit strip from the operator, and the strip is the part that goes to zero in a restructuring. That matters because in a bankruptcy the tenant can reject the lease, and rejection damages on an over-rented lease are capped by statute and paid in whatever unsecured creditors recover, which is normally very little. How that plays out in a specific filing is a question for restructuring counsel, but the direction is not controversial.

Coverage does real work if you can get it at the right level. Corporate EBITDA is close to useless. What you want is property-level or unit-level EBITDAR against rent, and a covenant to keep delivering it quarterly for the life of the lease. Under 1.5x on flex industrial with a non-rated operator and a private equity sponsor, I'd treat the rent as temporary. Also read the assignment clause hard. Sponsors sell portfolio companies, and a lease that lets an assignee inherit the guarantee without a net worth test hands you a worse tenant with no consent right.