RE: SAGD +Hmm - a serious question - you must be new here. As you can see, this board is only for disgruntled gamblers that bought a high risk stock and lost, and for dillusional pumpers who post over and ove the same baseless info.
To answer your question. Not sure how much you know about SAGD, but the steam chamber that develops around the injector well keeps expanding for years and years as it permeates the bitumen pay - so the solvent (which is gaseus when mixed with the steam) doesn't just hit the bitumen near the wellbore. If the wells are placed correctly and the pay somewhat homogenous, the majority of the bitumen will be hit with steam and solvent over a period of many years. The steam makes the bitumen flow and forms an emulsion of oil and water. The emulsion flows down the sides of the chamber, is collected at the producer well and (in CLL's case) is pumped back up the producer to the plant, where the water is separated. The ratio of steam and produced water to oil is key, and the solvent in the steam causes more oil to be recovered from the same well using less steam (so it will impact production, SOR, and reserves positively).
Steam is expensive from a capital equipment perspective and operations (the higher the SOR, the more water you have to process out). Nat Gas is cheap now, but at some point will cost more and thus a low SOR is golden in SAGD.
As for your second question about what happens to the 15%. It is a good one, and one that probably only Merle can answer. I suspect it is a combination of losses that add up to 15%. The NR indicates it stays in the reservoir, but unclear if they will eventually get more of it back, or whether it stays locked in the ground. From the NR, they have been using 10% by volume of solvent (which is a $#%$load of solvent when you scale it up to the whole plant - consider that ) and they are going to 15% by volume. But using 10% by volume they increased production by 23% and dropped the SOR by 15%. [As an aside - How does that work? X/Y=Z where X is Steam, Y is Oil, and Z = SOR. If you increase Y by 23% and keep X constant, Z should decrease by 18.7% not 15%. I guess the solvent (~10%) gets added to the steam injected so X increased by 10%. In that case Z would decrease by only 11%. Anyhow - would be nice to see the math on this stuff.] They suggest that SOR will be further improved once they reduce the pressure in the reservoir.
Bottom line, they were injecting about 10% by volume of solvent to get 23% more oil (bitumen) and of that 10% they are losing about 1.5%. Assume solvent costs about 2.5 times the sale price of bitumen. If the starting SOR is 4, then they use about 4*.85*.015 = 5.1%*2.5=12.75% oil equivalent to recover 23% - payback of about 2:1. Not sure what the captial and operating costs would be to implement + cost of transporting solvent - but it would appear to be high value. By increasing solvent to 15% they should get even more recovery.
Don't discount the impact of proven reserves. If it looks like they can sustain recovery of +20% for existing wells, then that could significantly increase reserves.