Even if you think that X is ok it is helpful to post both input shaper graphs.
The Z spike here might result from a twisting print head, depending on where the accelerometer is attached to or maybe a running fan.
However you have three different documents in this shaper graph calculation that might introduce different issues each, although they are all created on the same day.
So for your next IS measuring session please delete all old logs or move them to a different folder upfront.
How big is your printable area?
The spike at around 47 Hz is a bit low though in almost perfect shape…
Did you check your printers belt tension? How did you do that?