I did a search using Matsuka, and found this article - Performance Analysis of Photovoltaic Water Heating System
Not sure if you were referring to this.

The difference that their report and my system, is that they used 2kWatts of D.C. a as a base point with out any temperature controller, with a number of different immersion heater elements, and then switched between them according to the sun radiance, to maintain MPPT (Maxim Power Point Tracking).
My system boosts a 1.3kW array to 300v D.C. and then with the MPPT, it varies the voltage to the existing mains element, where the existing geyser elelent can be used to control the geyser temperature. Efficiency of the thermostat is irrelevant, as the cost of energy is negligible if one does not take the initial cost as part of running costs.

I have had the evacuated tube for 10 years, and compared it to the PV, there is no comparison, PV beats it hands down. Granted I require more roof space for the PV panels, so efficiency of the systems in comparison can only be done with respect to the amount of energy you can extract for the amount of surface area you have available.

If you have plenty of roof space, use PV, the installation costs are low as compared to the evacuated or plate heaters, in which plumbing and circulating pumps have to be installed. PV requires the frame bolted to the roof rafters, and then an electrical connection to the controller to the geyser.
With some observations done, I would rate the flat plate solar heaters better than the evacuated tube heaters, but again, to maintain the balance of summer over heating of the water due to higher sun radiance and lower hot water usage, require mains back up in winter to maintain the necessary hot water requirements. Actual temperature measurements of flat plate as recommended by installers for hot water usage, only yields a maximum of 48°C in winter at the end of the day if no hot water is used during the day.

So as I have stated before, when efficiency is touted about, please request the exact reference of the reference to the efficiency, and then make an informed decision.

I remember showing the Usedasun System to an ESKOM representative, and the first question he asked "So what is the efficiency of the system?" to which I answered "100%"
He said "Impossible!", to which i replied "What do you mean by efficiency?"
He replied "PV panels are only 15% efficient!"
To which I replied, "Yes, the efficiency refers to the amount of surface area that the sunlight strikes to produce the amount of equivalent electrical energy. Since I am not paying for the sunlight, so who cares, I simply add a few more panels for my application , after all I have plenty of unused surface area which the sunlight covers, and by using the PV panels I reduce the incident heating in the roof where the PV panels are. So efficiency is irrelevant"

To add to the mix, what the ROI of any system is not purely based on the installation and hardware cost, but also the cost of the back up mains to maintain the supply of heated water over the lifespan of the system. Therefor calculating the cost of a PV based system must include the reduction of the cost of the system if a flat plate heater was used, and then also subtract the cost of mains to maintain the heated water in the flat plate application against the minimal cost of mains in the PV system.

What is interesting is figure 13 of the paper mentioned above. Look at the performance of the MPPT output versus the flat plat performance. It seems that the values contradict the conclusion, if surface area is not an issue.
PV panels are getting more efficient and cheaper. There are now offers for 400W PV panels, so effectively 3 of these panels could meet your PV heating needs.