I like that the time base accuracy (1ppm initially) is as good as that of the SDS5000X and the channel skew specification is even a bit better (<100ps vs. <150ps). I like a lot that they even specified the trigger jitter for a scope of this price range and the value for CH1..CH4 is actually a magnitude better (<10ps vs. <100ps) than that from the SDS5000X. In both cases I assume that the (older) SDS5000X is just a bit less optimistic, but still.
It's also good to see that the signal generator is internal now (even though it's more of a toy with 3V amplitude at 16kPts for arbitrary signals). In sum, it really looks like a nice scope and not much of a step down from the SDS5000X other than the sample rate, maximum bandwidth and probe interface (but I really like that they kept the probe detection).
BTW: in the table "Models and Key Specifications", the datasheet indicates that only the SDS2354X can be upgraded to 500MHz and doesn't mention bandwidth upgrades for the lower models. Yet later under "Optional Accessories" in lists bandwidth upgrades options for all models.
SDSQOOOXP-4BW02 - 100 MHZ to 200 MHZ bandwidth upgrade (4-ch model) (software
SDSZOOOXP-4BW03 - 200 MHZ to 350 MHZ bandwidth upgrade (4-ch model) (software)
SDSZOOOXP-4BW05 - 350 MHZ to 500 MHZ bandwidth upgrade (4-ch model) (software)
SDSZOOOXP-ZBW03 - 100 MHZ to 350 MHZ bandwidth upgrade (2-ch model) (software)
So letting aside that the 2 channel model seems to be limited to 200MHz: is it sure that all 4 channel models can be upgraded to 500MHz?
imho internal generator is good option if remember that also SDS2000XPlus have quite good SFRA. Without external generator it can do SFRA from 10Hz to 50MHz. And this BodePlotII IS really good, like reference to others (when look this this class of equipments).
How you look 2 channel model is limnited to 200MHz when there is 2 channel upgrade to 350MHz listed.
But then with 500MHz upgrade. It need tightly understand and take to account what is told in data sheet.
First up to 350MHz versions time scale is 1000s - 1ns/div and with 500MHz upgrade 500ps/div - 1000s/div
But then it is also clearly stated: In interleaving mode bandwidth is 500MHz, risetime is 0.8ns; in non-interleaving mode bandwidth is 350MHz, risetime is 1ns
Interleaving mode=for one ADC chip only 1 channel is active (2GSa/s).
My opinion why 500MHz risetime is not 0.7ns is that BW fresponse shape also change. (different filtering? As we well know, scopes what have "max flat - brick wall" type response it also, of course, produce lower risetime vs gaussian fresp..)