How badly could someone be injured by concentrated sunlight? Recently-ish, I stumbled across an interesting short story (by way of Science Fiction & Fantasy Stack Exchange) where a soccer referee is apparently incinerated by concentrated sunlight.

Where the referee had been standing, there was a small, smoldering heap, from which a thin column of smoke curled up into the still air.

This is accomplished in-story by some 50 000 reflective tinfoil program covers, each about the size of a tabloid sheet.
What got me interested (physics is one of my peripheral interests) is the feasibility of this method in the real world. Wikipedia says a tabloid sheet is $279\;\mathrm{mm} \cdot 432\;\mathrm{mm} = 120528\;\mathrm{mm}^2 \approx 0.121\;\mathrm{m}^2$, giving fifty thousand people with a program each a total of $6026.4\;\mathrm{m}^2$ to work with. (Adjusting for less-than-perfect aim, I'd say closer to about $5000\;\mathrm{m}^2$.) If you could redirect that much sunlight at about a person's surface area, how much power would that be? What damage could you cause?

He couldn't have felt much; it was as if he had been dropped into a blast furnace...

Could that amount of power actually incinerate someone?

It occurs to me that the Wikipedia summary of the story (quoted on SciFi.SE) states only that the referee "collapsed and died". If incineration isn't possible, could the energy involved still be lethal by other means?
 A: A lens can burn badly at the focal point, one can easily start a fire and get a bad burn if one is stupid enough to focus for long on skin.
A $5 \ {\rm cm}$ diameter lens concentrates the power from about $2 \times 10^{-3}$ meter square. Taking the conservative $750 \ W/m^2$ (it can be $1200$ in my area) a  power of $750\times (2 \times 10^{-3})$  on $2 \rm mm^2$, or $1.5 \ W$ on $2 \ \rm mm^2$ of skin or paper or wood, sets fire and burns flammables. 
If the cross-section area of a man is $0.85 \ \rm m^2$ as Chris estimates, to burn with the power similar to a lens such an area , one should send $\sim 4.3\times 10^5$ watts on that area ($0.85/2\times 10^{-6}$). 
So if megawatts were concentrated on the man, depending on the time, first his clothes would burn, then the skin would dry up and burn.  Chris' solution gives megawatts (8.3) to cover this estimate, and I suspect the authors used the lens example to get at realistic numbers and needed 50,000 people in the stadium sun side..
About Archimedes setting ships on fire:

A test of the Archimedes heat ray was carried out in 1973 by the Greek scientist Ioannis Sakkas. The experiment took place at the Skaramagas naval base outside Athens. On this occasion 70 mirrors were used, each with a copper coating and a size of around five by three feet (1.5 by 1 m). The mirrors were pointed at a plywood mock-up of a Roman warship at a distance of around 160 feet (50 m). When the mirrors were focused accurately, the ship burst into flames within a few seconds. The plywood ship had a coating of tar paint, which may have aided combustion. A coating of tar would have been commonplace on ships in the classical era. 

Did myth busters use tar coated ships? Tar was the only water resistant material known to the ancients for waterproofing boats and ships. In addition the Mediterranean has very dry sunny days most of the summer, the watts/metersquare can reach 1200, ( good for solar panels)
Moral: Do not believe whatever myth busters say uncritically.
A: There's a solar furnace in France that can melt steel.   I don't know the collection area of it, but it's probably larger than 5000 m${}^2$.  Tin foil tablets ... I don't know.  But it sounds like the sci fi story is not outlandish.
A: At the radius of the earth, the solar irradiance is approximately $1.412\;\mathrm{kW/m^2}$, giving a total power hitting the foil sheet (assuming normal incidence) of $\sim7.06\;\mathrm{MW}$. The average human in America is around $1.7\;\mathrm{m}$ tall, and somewhere around $0.5\;\mathrm{m}$ wide, making his cross sectional area around $0.85\;\mathrm{m^2}$. This means that the human will have an irradiance of around $8.3\;\mathrm{MW/m^2}$ (again assuming normal incidence, increasing the angle of the foil with respect to the sun will decrease total power, and hitting the man from any angle will increase the power he feels locally). Now, this is where it gets tricky. The blackbody spectrum of the sun peaks in the visible, and is absorbed strongly by the atmosphere in the UV and decays in the infrared.  Tissue ablation requires somewhere on the order of $2\;\mathrm{MW/m^2}$ for an infrared diode laser peaking at 1470nm (see "Optical Sealing and Cutting of Blood Vessels Using Near-Infrared Laser Radiation." by Sarah Rosenbury). Now, that is from all the power directed at an absorption peak of the water absorption spectrum. Thus, because the absorption of human tissue in the visible is predominantly due to melanin and blood (hemoglobin-Hb), scattering will occur before absorption (except in the hemoglobin and melanin molecules) and more power will be lost. Not to forget mentioning that the penetration depth (how deep the light goes) is at best a couple of millimeters. Thus, while you are likely to cause burns, light his clothing on fire, and be generally painful, the damage would only be skin deep.  In order to penetrate more deeply one would need to use gamma rays (which is difficult due to atmospheric absorption) or something with a longer wavelength like microwaves (but they have very low energy per particle so it would require MUCH MUCH more). So as a final answer to your question, no. It is not possible to instantly vaporize someone under those conditions, let alone incinerate them.
A: On your figures, it's going to depend on conditions, and also critically on what the referee was wearing; but in general it would certainly put the hapless referee in very dangerous position. It probably wouldn't be quite so dramatic as in the tale.
On a typical day, let's say the Sun delivers $750{\rm\;W\;m^{-2}}$ intensity when straight overhead. Scale it by $\cos(\pi/3)= 1/2$ for a typical latitude. Scale it again by $1/2$ to account for the fact that the foils must be tilted roughly 45 degrees to aim the light. And add a further factor for random losses, so we assume that a tenth of the sunlight reaches its target. You still have $500 \times 750{\rm\;W}$ reaching the hapless fellow: that's $350{\rm\;kW}$. Let's assume a body is mostly water: a 70kg body therefore has a heat capacity of $280{\rm\;kW\;K^{-1}}$. So that heat loading, if all absorbed by the fellow's body, would lead to a temperature rise of a couple of degrees celsius a second. That's not good. As you can see, if some of my scale factors were increased for more "favourable" (to the killers) values, we might get a third of the light trained on his body: six degree celsius rise per second. That would very swiftly be lethal.
Now we come to what the fellow were wearing. And also his skin colour. his will be critical. If he were clad in a highly reflective spacesuit, he might fare pretty well. If he were wearing white and were quick witted, he might have the time to duck down and crawl away. If what he were wearing were wool or fire retardant, even better.
You would need a bit more light than that arising from the conditions in the tale to get a dramatic "vapourisation", but it would not be good for the referee.
It's interesting to compare my answer and Chris's as I've just looked at heat loading whereas Chris is looking at more specific safety limits on optical intensity for skin. The two answers roughly agree in their conclusions, although I should think that "skin deep damage" to a large fraction of the body could well be lethal, as any clinician or nurse in a burns unit will tell you.
A: Read about a solar crematorium which shows what is possible with 4 sq m of concentrated sunlight.
http://www.solare-bruecke.org/infoartikel/Papers_%20from_SCI_Conference_2006/22_wolfgang_scheffler.pdf
