The Data from a UK Boeing operator indicated that NPAs were causing the less stable of the approaches. This was addressed in two ways. An opc profile that covered NPA as a loft amongst other scenarios. Also training was adjusted to reinforce the practice of setting up (level and appropiately configured segments until the final descent fix) and flying the npa as if the wx was at minima (on ck days) with the pf head down until a pre agreed point. The transition to visual was also emphasied. This has gone along way to improve things. The tendency was for the pf to be in a kind of no mans land between viz and imc proceedures anticipating the breakout or effectively flying a visual thru the mcp, eg using the papis to adjust the v/s or vice versa, or not transitioning positvely to the visual segment either during practice or for real. The addressing of the situation is making huge improvements in Data.
They didnt need to issue a memo and the pilots didnt take it personally.