For scale and simplicity PV is hard to beat. Aside from efficiency considerations (below), complexity and reliability of thermal (e.g. steam) systems will be a challenge. It’s hard to beat a system with no moving parts!
As a rule of thumb, for a good solar location, assume 1000 W solar radiation per sq meter at the surface around noon in summer (eg Sydney Australia). A good PV panel will convert 19-20% of that to electricity, so around 200 W per sqm. A fresnel lens - which is most likely plastic, might capture and concentrate 90% of that energy providing around 900 W thermal power for the same area. This then needs to be transferred to a fluid for example to heat water or for another use such as power generation. The efficiency of this process could be (very) low, depending on the heat exchanger design (material, pumping, radiative losses etc). Converting heat to electric power is thermodynamically inefficient unless you can achieve very high temperatures (even modern open-cycle gas turbines struggle to achieve better than 35%). At the sort of fluid temperatures you could achieve with a fresnel lens any heat engine would be woefully inefficient - say 10%. So the 900 W of thermal power might provide 90 W electric power - at much greater complexity and cost.
Modern solar thermal power stations use mirrors to concentrate sunlight, but even highly reflective parabolic troughs can only achieve fluid temperatures around 450 C (it’s also the maximum temperature that the oil can withstand) which limits the efficiency to about 20%. Using heliostats (almost flat mirrors) to focus sunlight to a tower can achieve much higher temperatures and therefore thermal efficiencies, but these come with other challenges such as the need for stable, strong materials to handle the temperatures.