I guess it depends on what they actually mean by "efficiency" and how they are calculating their numbers. There are lots of ways to do that and they appear to be cherry picking the numbers.
1. The lenses are concentrating solar radiation at the upper lens panel onto smaller point destinations on the lower cell panel. So more incident radiation available per mm² of actual cell area. In that sense the cells will be more "efficient" in that they will produce more Watts that a similar cell that doesn't have a collector lens above it.
I'd guess that's how they get from 22% to 36%
2. They are using fewer cells per m² because they need to be spaced apart for this to
work. If they can produce the system more cheaply with the cost of the
tracking lens panel more than offset by the reduced number of cells per square metre, then it will be more
cost efficient to produce per m².
3. Since the total effective collection area is the area swept by the moving lenses, not just the surface area of the lens panel, they are collecting more radiation - so it can be a bit claimed to be a bit more efficient in in terms of the W/m² of area of the lower panel
So they can certainly achieve increased efficiencies by some metrics.
BUT:
They are the same cells currently in use. There is no change in the efficiency of the cells in terms of converting x Watts of incident radiation to y Watts of output
power. (The standard metric for cells) It will still be the same old 22%
The increased "efficiency" in terms of Watts per cell through increase "lensed" radiation must at least be offset by the decreased efficiency in terms of Watts per metre at the panel level because of the large amount of space between cells.
The system takes up more room because the lens panel and cell panel move relative to each other, increasing the surface area required per panel and this must at least offset the "efficiency" in terms of Watts per m² or Wh/m²/day of the "cell panel".
Bottom line, while individual cells may well generate more Wh/cell/day, and they may be a bit
cheap to produce per m² or per Wh , they will NOT be more "efficient" in terms of either the standard cell metric of conversion efficiency or the standard panel metrics of W/m² or Wh/m²/day where m² is the area occupied by the total system.
For a
boat with the usual demands of energy and space, those Wh/m2/day is the critical "efficiency" factor and this system won't really change that.