The flipper isn't really a full sdr though, it just has a very minimalist RF IC that has almost non-existent bandwidth.
For $400 you can get a limeSDR mini that can read and write 30MHz of spectrum at a time, ie the entire ham 70cm band all at once.
If you think a flipper is dangerous, plug in a dummy load and dump noise on L1 then watch your phones GPS stop working, or alternatively decide it's on another continent.
I think it's a case of good marketing and good packaging. Realistically you need a laptop to do serious field stuff with a flipper, but only if you plan on doing any testing and reconfiguration and only initially. The flipper isn't much bigger than the limesdr card, but it's nicely packaged and portable so once you have it ready to go you can throw it in a pocket.
The community also helps. The flipper is wildly overpriced for being a glorified happy meal toy but millions of people squeezing every ounce of functional potential out of a happy meal toy is better than a few dozen people writing academic papers with mostly high end industrial (cellular base station) and military applications.
Depends very broadly on your area of interest, but to throw out some random numbers and thoughts:
If you want to move data between two points, 30MHz of "bandwidth", depending on noise and signal, can be on the order of 30MB/s data rates or more assuming you're good at doing QAM or similar modulation. That's 50x what the CC1101 in the flipper maxes out at
If you want to search for a particular signal of interest (ie why does turning on my LED lamp open my garage door), that's more spectrum you can view at once, about 3x wider than what an RTL-SDR can receive. Similarly, you can view the entirety of a 30MHz wide emission as opposed to only seeing pieces of it.
You could monitor two different narrow bandwidth signal sources that are within ~30MHz of each other simultaneously, ie the 101.5FM broadcast channel and 121.5 airband guard channel. This provides the capabilities of something like a police radio scanner, covering the entire VHF or UHF land mobile band but without having to stop listening to find another signal and the ability to record the entire spectrum capture to disk so you can review all concurrent transmissions separately at a later time.
Above is a spectrum plot of an FM broadcast station using wide FM modulation and with some digital sub carriers on either side for song info etc. Other stations will be to the left and right of it and the "bandwidth" of the receiver determines how wide the plot you can view is.
Decoding trunked protocols frequently involves simultaneously listening to the control channel and data channels. If you only have access to low-bandwidth receivers, you'll need multiple, which gets into time sync problems.
Why would you need such a stack?
Article is analyzing unidirectional fobs, HackRF is half duplex so you could easily capture and analyze and/or replay the signal.
Only additional thing you need is a PC.
One thing to consider is that the payload will be encrypted so you wont be really able to tell apart what is the rolling code.
Hopefully fobs have stronger encryption so collecting enough sniffs and analyzing is insufficient (looking at tesla with their 64bit encryption, hopefully they upgraded).
Honda replay myth mentioned in the article is BS, it was popularized by ppl faking a simple replay attack while doing a more complicated one. If you record the fob command and the car never receives it, of course you can immidiatly after replay it to the car and car will accept it since RC is valid. But if you're sniffing while car is receiving it, RC gets updated.
If Honda didn't have RC, it would have been far worse than the KIA boys (overriding immobilizer protection and hotwiring the car) issue that did a lot of damage to KIA in US.
Yes, but a "HackRF clone, plus a Proxmark3, plus IR, plus whatever" probably isn't.