Yes, it is more efficient than just tealights, though still a very small source of heat.
You note that "all heaters are 100% efficient." This is partially true - all heaters convert virtually all input energy into heat, thus achieving near 100% efficiency. However, if that heat simply rises to the ceiling by convection it doesn't do much to warm the room or its occupants. Radiant heat flows directly from the source to people and objects in the room, and is much more effective as noted in Wikipedia:
The internal air temperature for radiant heated buildings may be lower than for a conventionally heated building to achieve the same level of body comfort, when adjusted so the perceived temperature is actually the same.
This system appears to be a more effective way to heat than an open candle flame, because the flame heats the clay pot, which radiates the heat to the room. The total heat input is still very small, however, as noted below.
The design can be viewed as a much smaller and simplified version of a masonry fireplace. A normal fireplace is very inefficient because most of the heat is carried up the chimney by convection. As noted here,
At best, an open fireplace is no more than 20 percent efficient.
Efficiencies can actually be much lower, and even negative if large amounts of room air are drawn out through the chimney.
An open candle operates similarly when viewed as a heater. This reference determined the amount of energy radiated from a candle through laboratory testing:
The radiative fraction was determined by finding the ratio of the radiative emission and m x Hc [my note: this is the total heat produced by combustion], which yielded a value of 0.17 +/- 0.01.
Thus, 17% of the heat produced by an open candle flame is radiated to the surroundings, with the remaining 83% carried away by convection. This convective loss is equivalent to the convective loss up the chimney for an open fireplace, as it rises rapidly to the ceiling where it is effectively lost.
A much improved design for a fireplace forces the combustion gases to flow through a circuitous path made of masonry materials, which absorb heat from the gases and radiate it to the room. This improves the overall efficiency of heating to much higher levels. Here is a study which showed one masonry fireplace achieved a heating efficiency of just under 80%. The stove looks like this in cross section:
The gas flows are similar to the flower pot arrangement as shown here:
Thus, the tealight and flower pot function as a small-scale masonry fireplace, reducing the convective losses and increasing radiation to the room, as opposed to an open candle which functions like an open fireplace, with most of the heat lost by convection up the chimney.
While the flower pot does increase the effectiveness of the candles as heaters, a few calculations can show how little heat it actually produces:
Tea lights are made of paraffin wax, and are commonly about 38 mm in diameter and 16 mm high, and burn for 3 to 5 hours (per Wikipedia). Paraffin wax has a density of about 900 kg/m3, and a heat of combustion of about 46 MJ/kg. running through the math, we get 0.75 MJ/candle, and assuming a 4 hour burn time and 100% combustion efficiency (which is close, though not exact) that is a power output of 187,500 J/hr, or 52 watts.
If 4 candles are burning at one time, this is 208 watts - probably quite close to the heat being generated by the two computers and two desk lamps in the room shown in the video.
Would it be better to put a light bulb inside a flower pot? Probably, since it wouldn't produce any combustion by-products such as carbon monoxide, which flames inevitably do. But from a purely economic point of view, it may actually be cheaper to use the candles, if the pricing claim is true. If the candles truly cost 1 £ for 100, then the heat costs about 5 pence per kW-hr, which is lower than the 14 pence I found by looking at British electric rates. However, the cheapest source I found for these candles in the US is $6 per 100 (3.75£/100). At this price, the candle heat costs about 18 pence per kW-hr, which is more than electricity.
Finally, there is nothing new about this idea. Googling "Candle Heater" brings up all sorts of sites, such as this one with flower pots