1 square meter at Earth sun distance has roughly 1366 watts.
1 square km has 1366 million watts.
10 km square has 136,600 million watts
100 km square has 13,660,000 million watts
13,660,000 million watts equals 1.66 10^13 W
With solar panels capturing 100 km square of solar energy with 10% efficiency the solar panels would generate 1.66 10^12 W of electricity

Earthlings consume 1.5 x 10^13 W
Includes transportation and energy from all fossil fuel used.
http://en.wikipedia.org/wiki/World_energy_resources_and_consumption

In terms of expressing as kw/h per year: 1.66 10^12 W of electricity
is 1.66 10^12 W times 8760 divided by 1000 or 1.454^13 kw/h per year
or 14.54 trillion Kw/h per year

In terms of just electrical energy Earthlings consume:

"17.3 trillion kilowatthours in 2005"
http://www.eia.doe.gov/oiaf/ieo/electricity.html
"Worldwide net electricity consumption in 2001 was estimated to be 13.9 trillion kilowatt-hours.
If global electricity demand grows at the average annual rate of 2.4 percent predicted by the EIA,
24.7 trillion kilowatt-hours would be consumed in 2025"

So, in years 2030 and beyond global demand for electrical could easily exceed 30 trillion Kw/h per year.
And 100 km sq area in space could provide close to 1/2 of that electrical energy needed beyond the year 2030 or provide over 2/3rd of our current global electrical consumption.

Now, let's side aside for the moment the technical challenges involved, and suppose one had some company which simply had 100 km square area of solar panel and the company wasn't going to try to ship that power down to the earth surface, but was willing to sell electrical power at 1 cent per kw/h to any party wanted to buy it at the location of the solar panels. So at that price, if this company could sell all the power generated: 14.54 trillion Kw/h per year they would a receive yearly income of 145.4 billion per year.
If someone wanted to buy all this energy per year, and would lose 75% of the energy in transmission losses- buy 4 and deliver 1. They would need to sell the power for more than 4 cents per kw/h or the
Earth surface- 4 cents plus cost of capital and profit.
If they sold the electrical power on Earth for 5 cents per kw/h. The yearly gross would about 29 billion.

This roughly means if they spend about $90 billion in capitial, and they had a low operating cost and could buy all the power and sell all the power for 5 cents per kw/h, they would be profitable.
Suppose if they could spend more in capital cost, but get only a 50% loss of energy in transmission to the earth's surface- buy 4 deliver 2. That would make their year gross about 58 billion. Allowing about
$150 billion spent in capital and still getting a profit. And reducing the transmission loses further also allows more capital to be spent to achieve this better efficiency.

It should noted that if you had a space elevator this could allow you send power thru power cables- and this might one way to increase the efficiency of the power transmission. So if the space elevator
only cost 50 to 100 billion dollars to build it could be part of a solution of lowering transmission costs.

Now, let's return to the 100 km sq solar panels, if all the power generated can be sold at 1 cent per kw/h, then it has 145.4 billion per year in revenue, which means it's going have to cost somewhere
less than 1/2 trillion dollars by the time it's finished being built and delivering it's full power.
So, how much would 100 km square of solar panel weigh?
It's 10 billion sq meters.

And: "The use of ultra-thin (less than 50 micrometer) / low mass (0.2 kg per square meter) photovoltaic cells with a high specific power value (up to 200 W per square meter) minimized the mass of these vast arrays." http://www.astronautix.com/craft/mars1989.htm
So, let's say 1 kg per sq meter as about the minimum one could expect, though perhaps 1/2 kg per sq meter is possible.
So, 10 billion kg in total weight, and need less than $50 kg to make and ship it there.
Can't ship it there from earth using chemical rockets. And is even difficult to ship there from the moon using chemical rockets- the price of lunar water would probably need to somewhere around $1 per lb and power around 10 cent per Kw/h.
Obviously if shipped from the Moon you would need to be able to make solar panels on the Moon. And the price of solar panels on the moon would need to be fairly inexpensive.
On earth today, you can buy the cheapest solar panel for $3 per watt. The panels talking about here have 136 watts weighing 1 kg per sq meter so even if priced at $1 per watt without including the shipping cost it's way over the budget.

Hmm, so the cost of shipping from the moon, won't be as significant as the cost to make the solar panels.
The Moon would have advantages as far as making solar panels cheaply.

Continuing.
First you could made these solar panels on the moon in 100 by 100 meter sections, 1 kg per meter it's 10,000 kg. And 1/6 of this because the Moon's gravity, so 1667 kg.
You would still have a exactly a million pieces to put together but it would be easier.
You could tow these 100 meter sq sections with a couple of rockets.

Second, making pure silicon requires a vacuum. Moon has lots of
cheap near perfect vacuum. And with the really, really pure silicon you grow a silicon crystals, and then saw it up into thin wafers. Moon's low gravity might good for growing crystals. And on earth sometimes the vacuum pump fails and ruins a batch- you shouldn't have this problem on the Moon.

Some references with quotes regarding making silicon for solar panels:

http://www.solid-state.com/display_article/319150/5/none/none/Feat/Crystalline-silicon-solar-cell-manufacturing-requires-vacuum-based-solution
Title: Crystalline silicon solar cell manufacturing requires vacuum-based solutions

"The silicon produced via this process is called metallurgical grade silicon and is at least 98% pure.
Using this method, silicon carbide, SiC, can form. However, provided the amount of SiO2 is kept high,
silicon carbide may be eliminated, as explained by this equation:
2 SiC + SiO2 > 3 Si + 2 CO.
In 2005, metallurgical grade silicon cost about $ 0.77 per pound ($1.70/kg)."
http://www.minormetals.com/content/html/minormetals/about/AboutSilicon.htm

"making solar grade Silicon at a production cost of around $7-9 per kilogram, compared to a production
cost of $25-30 via a typical Siemens plant.

The cost of the Silicon in a current panel or cell is about $1.50 per Watt. RSI will reduce that to
about $.25/W over the next 5 years. This should help reduce the current retail cost from $4/W to
under $2/W, achieving grid parity at about $1/W, targeted for 2012-2015.

How RSI Achieves Low Price
Most Silicon plants make electronic grade "8 Nines" (99.999999%) purity needed for semiconductors, using the Siemens process. A National Renewable Energy Laboratory study found that solar Silicon
purity only needs to be six nines (99.9999% pure), with emphasis on Boron, Phosphorus, and select metals. RSI has created a new proprietary process that makes only solar-grade silicon of six nines
purity, without needing all the costly equipment and tri-chlorosilane gases used with metallurgical Silicon feedstock.

Thus, a typical 5000-ton plant can be built in 15 months, at one tenth the cost (only $50M) of Siemens process plants ($500M), and be easily ramped to 15,000 tons per year capacity."
http://pesn.com/2007/05/02/9500469_RSI_Silicon_wins_MIT_contest/

Ok, conclusions:

Nick will probably delete this.

And solar panel 100 km sq, would have serious problems related to solar wind.
Sue me.

The party with 100 km sq solar panel would probably need to sell power at solar panel location for more than 3 cents of Kw/h, instead of 1 cent per Kw/h.
Obviously, I don't think this could work within the next decade.
Point is this has to do with long term energy needs of Earth, so
perhaps as soon as 20 year in the future, though if radical steps
were taken such as offering zero taxes for 30 years to those involved in such enterprises, one could see it happen in less than 20 years.