O.K. Shows that the 0.5 dB drop at low frequency is appparently an instrument artefact, but it's basically measuring correctly up to 5 MHz.
So the 600 kHz cut-off in your measurements is a real hardware effect. I continue to assume that it's caused by a capacitance of about 25 pF parallel to your load, mainly the input capacitance of the measurement system and some additional circuit capacitance.
The issue has been already discussed in your previous thread.
https://www.edaboard.com/threads/341771/ I wonder why you rescheduled the question without considering the previous discussion.