Yes. 
I will only need to start U-Boot. I hope U-Boot has a boot delay. So I can enter the U-Boot command line.
So my first goal is, to get a working serial console.
EDIT:
U-Boot boot delay: I mean the delay before U-Boot loads the Kernel (Autoboot).
Fine I won't explain it.
Could you please go do somemore reading and enlighten yourself and when you understand what everyone else here is trying to say and explain it to us in a way that a hardware engineer will understand so that next time a hardware engineer asks this question we can all link this thread.
Because I think you can't explain it. No furthermore comments from me.
EDIT:
As I wrote: I work as hardware and software (low-level programmer) developer. Hope you understand me now.
That is so typically to all OpenSource Communities. When a external guy like me will contribute to something, then you get useless (stupid) comments from some developers...
I DON'T need to release any details for the RAX200 or add support for OpenWrt. Don't forget this (OR I will fork OpenWrt).
Sorry, I'm upset from some arrogant developers.
Well, enjoy this time vampire of a project, fingers X:ed.
Prove me wrong
Explain to me
Good luck with that. Broadcom doesn't just hand those out to random people: you have to be a proper corporation with a licensing deal to get access to those.
Interesting assesment of the "Open Source Communities"...seem like you've gotten the same reception before.
Maybe it's because of your approach.
Some very knowledgeable folks have given you accurate and detailed answers here.
It seems you have a high opinion of yourself, and your skill set.
I am a former software engineer, and have worked with hardware engineers on different projects.
You don't seem like either.
You also don't seem to understand basic concepts about SoCs, Broadcom limitations in OpenWrt, and how OpenWrt works in general.
The arrogance in this thread is coming from you.
It is your time to invest, and your time to waste.
Just don't waste other people's time, which could be better spent elsewhere.
You do you.
I love the arrogance here: you're basically saying "I'm better than any of you and I know better than any of you, therefore anything any of you say is irrelevant and wrong!"
Well, guess what? Many people here also do or have done hardware- and software-development as well! You're not some unique magical unicorn. Also, many people here have worked on OpenWrt for years, maybe even decades -- they're not just some newbies with zero knowledge of what they're talking about.
Yes, if you can get full access to the Broadcom SDKs required to build an image suitable for running on the SDK, including Broadcom's proprietary WiFi-drivers, then that's great! Go for it. But it won't help OpenWrt none, since the project can't include those proprietary drivers.
Sorry for being offtopic, but dude !!! you're a software/hardware engineer and you buy a gaming router @450€ ???
Seriously, for that price you can get a X86 mini PC + AP, fully compatible with openwrt, and 10x faster...
Im not an engineer, a software developer, and i don't like to argue whit people whitout reasons.
Probably from your point of view, i shouldn't even reply to this thread, but it is soo tempting, i just have to, sorry 
So, i'm not an engineer, but i can tell you from my experience, whit companies like quectel qualcomm and all the illuminati team, that treat their software like holy graal, you won't go anywhere.
The reasons for this, is that no one will give you the hardware specs you need, and for sure they wont give you the driver source code you would need, to write the drivers to make things work.
Whitout the hardware specs, you cannot write a driver from scratch, because you didn't made that hardware.
And they won't give you nothing, unless you have millions to make a contract whit them.
And even in that case, you couldn't share your project, because it would not be under any GPL or any free opensource license.
The only way you could make something, is that, if you hack their server and get top secret code, just like the hackers that hacked nvidia.
But, at that point, OpenWrt could not receive the code, or broadcoam and qualcomm would issue OpenWrt for copyright violation.
Guys developers, ppl better than me, tell me honestly.
AMIRIGHT?
Cheers
Why would you replace the WRT32X that still has a fast CPU and near perfect OpenWrt support (mine has a ~160 day uptime on a 21.02 build from Dec with a swap to kernel 5.10 (easily hits 500/35 Mbits with SQM cake, 120 MB/s on Samba4, works with DSA, nftables, etc.) to junk like the RAX200 that has no open source support and will never have OpenWrt support? If you want a high speed network for your fancy Cat8 use an x64 2.5Gb mini PC, or at least something like a NanoPi 4S might have made sense, throw on a gigabit switch and wifi6 AP like the U6-Lite and your good.
For new wifi 6 routers only buy MediaTek or Qualcomm. That should be pinned at the top of the forum. They have open source support in progress and will likely run OpenWrt very well once we see kernel 5.15 for all targets and a 23.0x release (~1.5 years from now).
You can actually reverse engineer the proprietary blobs like it was done for the Panfrost driver for example but that's a long and tedious way to go