Your remarks are true in general, but do not apply in the case of these two models. So far I've tested ten 223B and six 223C models with consistently bad results on both Taiwanese and Indian-made Verbatim MCC 004. The units were bought from different vendors and tested by myself and 2 colleagues of mine on five different, clean, updated, non-overclocked systems running WinXP x32, WinXP x64, Vista x32, Vista x64 and Win7 x64. All burns were images fed from fast, defragmented hard disks and under mimimal system load conditions. Latest versions of Nero and ImgBurn were used in all cases. The discs themselves were taken from many different batches in order to eliminate the "bad disc batch" possibility. Out of all these units only one (a retail 223B) burns Verbatims well, and again not at x22 speed.
I feel that there has been a big quality control problem with these two models. Maybe sub-standard components were used, I really don't know. The latest Samsung x24 burners seem to be able to burn Verbs well at full x24 speed, which proves that Samsung techs have managed to bypass the problems that plagued some of their x22 variants. With 223B and C variants it seems to be luck of the draw: If you're lucky to have bought a non-affected unit you'll be happy. A lot of people out there are not so lucky.
BTW, my own retail 223B can consistenly burn with decent results up to x18 speed only. At x20 it starts dishing out coasters, but some burns are good. At x22 most burns are coasters with very rare exceptions. I always use MCC 004 or Yudens. If these models are incapable of burning at their rated x22 speed on some of the best media in the market, it's surely not the tester's fault!