A spreadsheet would work well here.
How?
The OP has given enough detail of his requirements to make a informed recommendation.
- no PC/USB type oscilloscope
- 70/100 MHz analog BW (upgradeability preferred)
- 4 analog channels (separate knobs per channel preferred, not like LeCroy - one set of knobs for all channels)
- upgradeability "MSO ready" to 16 digital channels usable simultaneously to the 4 analog channels
- serial decoding (and triggering) for analog and digital channels (I²C, RS232, SPI, ...)
- maybe CAN decoding and triggering upgradability
- maybe USB/2/3 decoding and triggering upgradability
- availability from european/german distributor (warranty)
Any manufacturer's website and instrument datasheet will show whether their product will meet his spec or not.
Simple.
There are always tradeoffs. Given the shopping list of features, the budget doesn't look hopeful.
So, instead of just stating 'gotta haves', rank them by priority and then compare actual scopes versus features versus price. Everything is negotiable when it comes to money. Otherwise we would be discussing real high end scopes.
An example is the MSO option. For true mixed signal, it is required, no question about it. As a logic analyzer, well, a real logic analyzer may have much more sophisticated triggering. Dave did a video on the topic of MSO versus logic analyzer. How many times does true mixed signal come up in the hobby community? If this scope is for a hobby, maybe never.
So, if mixed signal is a real requirement, fine! If it is just a line on a wish list, well, maybe there are better alternatives. At the least, it shouldn't drive an unnecessary solution.
Does knob/channel REALLY matter? What difference if you have to tap the channel select button before using a common knob? Scopes are available both ways but does it REALLY matter? When it comes to money?
Is sample memory size a feature worth pricing? It would seem to me, as a new user of DSOs, that more memory is a good thing. But is it the kind of feature that is worth money? To me, yes!
Only the OP can put worth to features. I don't imagine I will ever do true mixed signal that I can't view on 4 analog channels. The OP might! I don't object to hitting the channel select button but the OP might. Or he might not if there is money involved. Think about real scope usage: I tend to use just one channel more than half the time. I might use another channel as a trigger but once it is set, I will probably turn that channel off. So, once I select the active channel, I will never do it again. It will simply be the only active channel.
There are probably a dozen candidate scopes to consider (if we blow off USB) and the prices will range from a few hundred to a couple of thousand. How to choose?
If USB is REALLY a requirement, none of the scopes discussed above need apply. But a USB sniffer might be handy!
I just don't see how people make decisions without reviewing a ton of facts.