There may be problems from parts used with the isolation transformer. A transformer is just wires around a core, but isolation devices often come with a host of things inside the enclosure that houses the transformer. These different items may cause a lot of unintended problems. The voltage and overall current delivery might come within specifications, but that doesn't mean saturation during peaks won't occur, or that voltage and current are in phase.
Non-understandings of a lot of factors involved with such devices, including passive filters, cause an awful lot of issues. They give other similar but correctly designed things a bad name. I've been looking at the possibility of making special filter that can handle Boulder's biggest offerings, without causing any bad affects like seen here. It would not be an isolation transformer, and would be highly treated to prevent saturation problems.
If I had to take a guess the isolation transformer made the high frequencies sound a little muted, and things lost their general sparkle, you might say. I'd be interested to know what isolation transformer was used so I can research what is used in it that can cause the problems found here.
Anyway, for example, the Boulder itself uses 4x transformers, so transformers themselves are certainly not an evil, but rather something that can be implemented well or poorly.