I think you summarized it well. If they don't define properly what 34% improvement means, it's hard to take them at face value on the rest.

The article is too vague to understand what they mean by "the world's first-known quantum-enabled algorithmic trading with IBM"