Just quoting the article:
“from around 6% to over 27%, close to the limits of what single-layer photovoltaics can achieve today.”
But I guess it depends too on how many layers we’re talking about which isn’t specified.
You’re misrepresenting what they say:
During just five years of experimenting with our stacking or multi-junction approach, we have raised power conversion efficiency from around 6% to over 27%, close to the limits of what single-layer photovoltaics can achieve today.
We believe that, over time, this approach could enable the photovoltaic devices to achieve far greater efficiencies, exceeding 45%.
“6% to over 27%” isn’t the range of what they can achieve now. 6% efficiency was where the technology was at 5 years ago, and now they get 27%.