It sounds promising but to really know if you have the additional bandwidth you'll need to feed in a signal with an RF gen to verify the 3 dB point or measure the rise time of a known fast enough square wave.
I'm still investigating it but seems no true BW upgrade
hum, we may try a little further... I tested with my 8656A today...
The FFT is enabled
The vertical 25ns and 10ns makes no difference, frequency measurement doubles and quadruples
Can't see signals beyond 26Mhz the trigger gets funny, -3dB point is 25.3MHz, with firmware 1.1.5 for 1604B or 1602B, makes no difference.
So, its a half a hack, enables a near to useless FFT mode.
I believe the FPGA must be updated or maybe a resistor config in board.
One thing I've noticed at version 1.1.5 in both 1602B and 1604B, the measurement menus does not works properly after update. You cant select measurements, the scope works fine but the values wont "select", there is a workaround tough, to do a power cycle.
I've flashed the 1602B firmware back and open a ticket with Keysight, that's a serious bug