I don't think you can make linear price estimations because of the defective yield rate. If you are trying to cut 6 55" screens out of a sheet and only have to throw away one because of a defect, your cost per unit only slightly increases. If you are trying to make 2 77" screens and one is defective, you just doubled your cost. Yes, I know they would probably mitigate the lost by cutting them differently based on how they came out, but there has to be a premium placed the 65" and 77" because of the greater potential loss if the sheet is bad.
Of course what you are saying is correct (which is why I purposely used the term 'raw' in my post). On the other hand, if you simulate the yield, you will see that yield factors are secondary once the yield numbers reach 90% as LG is claiming on the 55".
Out of 100 sheets, there will be 600 raw 55" panels out of which 540 will be good, so 40 out of 100 panels will have a single defect.
3 65" panels uses about 71% of the area of 6 55" panels, so only about 28 of those defects will kill 1-out-of-3 65" panels and 100 sheets will produce about 272 good 65" panels (90.7%).
2 77"" panels uses about 65% of the area of 6 55" panels, so only about 26 of those defects will kill 1-out-of-2 77" panels and 100 sheets will produce about 174 good 77" panels (87%).
These number get worse quickly as 55" yields are brought down from the 90% level, which is why the only reason I can see the 77EC9800 being priced much above $10,000 is because LGs yields are far short of the 90% they have indicated and they don't really want to begin selling the 77EC9800 'for real' until yields stabilize at the 90% level (hence price prohibitively high to squelch demand).
Also, another factor is that 100% of the cost is not associated with the panel and most of the remaining costs (in electronics and such) is relatively fixed independently of screen size. If the OLED panel cost is only 75% of the TV cost, for example, 65" and 75" panels could be sold for even less at equivalent margin/profit.