To me, there is not a clear winner. There are positives with each choice, and I think you would figure out a process and be happy with either board. I have a Rambo, and I run headless for the same reason you want to run headless. If I bought it today, I’d probably get a SKR Pro. But if an SKR Pro was not going to be available in the near future, I’d wouldn’t wait.
I doubt you will see any quality difference in cutting between the two boards. In theory the SKR Pro is a bit better. It has hardware math support and runs faster meaning it is better at processing arc g-code commands (G2, G3). Most CAD and/or postprocessors will allow you to turn off arc commands and emulate the arc with a series of linear moves, and I doubt you will see any quality differences cutting wood with arcs turned off. And the Rambo is plenty fast enough to process linear moves at CNC routing speeds.
The SKR Pro (running Marlin) has an edge at laser engraving speeds. On the other hand, the firmware GRBL seems to handle speedy laser engraving better than Marlin, and there is a version of GRBL for the Rambo board. With XLoader, it is easy to swap between the two. I don’t know if that version of GRBL supports the LowRider. It does work for the MPCNC.
As for running headless, the touch display sold with the SKR Pro is easier to use when setting up jobs electronically. If squareness is important, I set up my jobs electronically using the electronic endstops. But, for most of my jobs, squareness is not critical, and I set those jobs up manually. For the LowRider, with the dual motors on the Z axis, electronic squaring is more important (I run a Primo). The Rambo display is just as easy for manual jobs.
The TFT display running in touch mode is running its own firmware and sending those commands across serial connections. We have seen a bug or two in the TFT firmware (now fixed), and serial connections using the handshake used by Marlin are suspectable to electrical noise and possibly other issues. But the TFT can also be run in Marlin mode. So, on the TFT, you can run your jobs in Marlin mode and get around any potential bugs and noise issues associated with touch mode.
On the forum, we see fewer hardware failures and fewer problems with the Rambo board. As some who likes to tinker with Arduino boards, I’ve seen my share of blue smoke and dead microprocessors. The Rambo boards survives problems (usually user mistakes), that would instantly kill other boards.
On the Rambo, the stepper-drivers are hardwired, meaning if one goes, you have to replace your board. On the SKR Pro, they are socketed. On the other hand, In the years I’ve been following this forum, I only remember one Rambo board that had (just) a dead driver, and everything else was working.
The TMC 2209 stepper drivers used on the SKR Pro are quieter, but that doesn’t mean anything with a router running.
I think the TFT sold with the SKR Pro will work with the Rambo board. I don’t see why not, and none of the bright people on this forum raised a red flag when I mentioned it before.
The is a hardware weakness with the SKR Pro related to endstops. We saw a flurry of problems at one point, and almost nothing after. I’m thinking it was that weakness combined with some out-of-spec components that resulted in the problem, and current boards are fine.
We have seen a (very) few SKR Pro boards lose their bootloader. An inexpensive hardware board was required to burn a new one, so it is fixable. I’ve never seen a Rambo board that had a bad bootloader.
It is easier to burn new firmware on the SKR Pro. Just drop the firmware on an SD card, and fire up the control board. The Rambo requires a USB cable and a computer.
The SKR Pro has the ability to add a module to allow you to communicate wirelessly, so that might be interesting. There is some setup involved, and, if I remember correctly, the connection is really slow wrt uploading g-code files. You would still want to sneakernet your g-code files.
—> So, the pros and cons are muddled. Buy either one, and I think you will be happy.