great things have been accomplished with rather terrible scopes and a little creativity
An excellent observation! Remember, what we scoff at as "junk" today would have been witchcraft even as recently as the 70's and 80's. Yet we were building microwave communication gear, landing men on the moon, etc. despite the "limitations" of the test equipment of those eras. Ever look into what they called "storage scopes" back then? Camera hoods on CRT scopes... hyper-complex CRT bottles with lots of weird grids, strange phosphors, etc. A minor miracle some of that stuff could even be physically manufactured, let alone in any volume, but that was the only option back when they didn't have ultrafast monolithic flash A/D's and ASIC/FPGA to handle the resulting bandwidth to memory - or even that much memory, at the necessary speeds!
If someone handed you a spec sheet for one of those scopes today, you'd dismiss it as worthless. Yet much of what we take for granted technologically today is BASED on things accomplished with such tools. That's why I'm slow to say that an older, or less capable, piece of R&D equipment has no value. Most of us don't usually work on projects where we need, or can tell, the difference between an 8-bit and a 10-bit A/D scope... or a 0.001% DMM... or a power supply with under 50uV of noise.
I have an Agilent Cell Test unit, a sort of all-in-one box that includes a scope, spectrum analyzer, sweep signal generator, demodulator, etc. Everything tops out at about 1GHz because it was meant for the days of analog cell service. Its original price was $40-60K depending upon configuration, it was state of the art, and my HP buddies tell me they sold thousands of them. Today its specs are dated but guess what - physics hasn't changed so it can do just as good a job today as when it was SOTA. I'm not tossing it out just because there's something better out there.
Back in the late 70's while in high school I paid $1400 for my Phillips PM3214 25MHz dual trace delayed timebase scope. Thanks to inflation a dollar then was worth a whole lot more than a dollar today, and yet today the OP purchased twice the bandwidth, twice the channels, with storage and memory, far more flexible and better triggering, probably 2/3rds less weight and volume, for about one-fifth the numeric dollars and even fewer adjusted dollars. And 50MHz probably covers 75% of the projects discussed on this site! Freakin' remarkable.
What I'm trying to say is "terrible" is a relative term. "Terrible" today was the bleeding edge a few years ago!