A bit of background from some earlier articles I've seen added to this;
The concept is based on IBM technology aimed mostly at thermal management in coimputer servers, but with applications to solar power.
See e.g.,: https://redd.it/2afct9
Traditional PV has the problem that elecrical generation is degraded at higher temperatures. This is especially the case where sunlight is concentrated on the collector. That's why most PV are flat panels rather than arrays of (cheap) mirrors + panels. Another factor is that flat panels don't need to track the sun (you point them generally south and angled by your latitude), hugely simplifying installations.
The "5,000 suns" bit here basically means that a total reflective surface of 5,000x the PV area is used. This also means that heat is approximately 5,000x greater. An advantage is that you need far less of the expensive PV collector. A problem such designs is encountering is that PV really isn't that expensive anymore. Oh well.... Of problems to have, this is a relatively good one.
Getting both electricity and heat from the same installation, if you've got something to do with both, can be useful. Though it often isn't. As the images accompanying the article show, the "sunflower" design ... isn't exactly unobtrusive. Where panels can be designed directly into new or existing construction, studding your walls and roof with heleostat parabolic dishes is ... more challenging.
There are also integrated solar PV + thermal systems. As others have noted they increase complexity and installation costs, though if you can stomach those the benefits in terms of overall fuel and mains savings are substantial.
https://en.wikipedia.org/wiki/Photovoltaic_thermal_hybrid_so...
A few further points:
● The real key in solar isn't efficiency but cost. Single-layer solar is limited to about 37% efficiency because physics, the peak theoretical efficiency is about 85%, at high cost, again, because of physics. You're starting from a ~1kW/m² flux, with reductions from there for panel efficiency, spacing factors, capacity factor, inverter (direct DC use would be a 10% gain), and net storage costs. The 85% efficiency gains of the Sunflower would all but certainly be eaten up through installation, maintenance, and manufacture costs.
● Solar and thermal can be combined, and with sufficient thermal mass, gains are fairly substantial.
● Total collector area matters. A small number of high-effieciency collectors ... collects a small amount of energy. A large, cheap area of modestly lower-efficiency collectors beats the efficiency gains. It's usually possible to increase collector area 2-10x. It's not possible to increase collector efficiency 10x starting at a 20% baseline.
For some massively impressive examples of what can be done with net-zero construction in Fairbanks, Alaska, see Thorsten Chlupp's Reina LLP:
http://www.reina-llc.com/
I particularly recommend the video series.
http://www.reina-llc.com/resources/videos1/
The following two videos are long (2+ hours) but incredibly detailed. Particularly impressive are Chlupp's use of thermal mass (both static insulation and a central thermal-stratification 5,000 gallon tank), and the consequent challenges (air exchange, heat exchange on all interfaces, including freshwater and sewerage), and most critically: moisture control -- condensation across a 100+ degree thermal gradient is a massive challenge.
If you're looking for solid construction details, these are an excellent resource.
Path to Net Zero Energy Series -- Alaska's first Net Zero Homes
https://www.youtube.com/watch?v=AtHkvpRI6fc
Alaska's First Net Zero Energy Homes Performance Update
https://www.youtube.com/watch?v=Xen_VWyDezY
2nd video is an update on the home described in the first. Goal achieved, except for plug-in Prius accounting.