[peteforde]

I understand that you're trying to flex a bit to prove your point, but a 2-lead rectangular discrete and dead-bugging a QFN-24/28 are in entirely different spectrums of difficulty.

Edit: Apologies, in my head I was replying to a different thread. I would still say that if you're working at that scale you should deploy appropriate tools to make the job simple, fast and repeatable.

Doing work on extremely small parts without a stereo microscope offers extremely small returns once you've finished proving your point to nobody watching.