I shoot with the Standard Picture Control. I rarely shoot NEF + JPG fine these days but rather NEF only. The sharpening of the Standard PC helps make a clearer LCD image for previewing. Turning off all sharpening in camera makes for a mushy image on the LCD. In post, usually CaptureNX-D these days, one of the first things I do is turn the sharpening all the way down. I do whatever I'm going to do in CNX-D or occasionally CNX2 and then pass the results on to Photoshop in a 16bit TIF. I specifically do not want this TIF sharpened.
In Photoshop I'll do additional image processing and save the TIF as a PSD file. If I do sharpening with the full sized image it will be a stamp of all visible layers that can be turn off or discarded of later. It will usually be the top or next to the top layer in the PSD.
When down sampling I create a new flat PSD file and then duplicate the background layer. This new layer will be used for sharpening. I frequently down sample in stages. I normally sharpen lightly before down sampling to retain fine details. This can be as light as 5%, 0.3 pixels, remove lens blur. I'll frequently down sample 50% at a time. If I have troublesome diagonal lines I'll down sample in smaller steps. Frequently this isn't needed. I can erase though areas of the sharping layer if they are troublesome. I can back off the opacity of the sharpened layer if I've sharpened to much, e.g. 100% down to 70%~50%. I use basic smart sharpening a lot. I sometimes use the unsharp mask and sometimes high pass sharpening. I may mix all three.
What I do not want is to sharpen the TIF I send to Photoshop and I don't want sharpening of my Photoshop PSD until very late in the PS work so I can throw it out if I need to. The final size and sharpening of an image is tailored for the specific use.
I'm still learning. Anyway that's my 2 cents on the subject.
Dave