@Ciaran_Whelan I wouldn’t recommend any smoothie-based board for laser engraving. But, if you had to have Smoothie, an AZTeeg is preferable to a Cohesion 3D.
Personally, I prefer TinyG/G2Core, and you can run that on a 328-based Arduino (TinyG) or Due (G2Core).
The Due is preferred as it is 32-bit, faster, more memory, better functions, but there’s also the Zero (I haven’t touched on, let alone tried). For the Due and Zero you will need a 3.3V-friendly CNC shield.
The grblShield from TinyG people works fine, but there are others on Aliexpress, from Protoneer in NZ, etc.
You can run your spindle and laser on the same controller - use the ‘coolant’ pin to turn the laser on and off, implement PWM/TTL if your laser controller supports it (the jtech certainly does). I’m in the middle of a tech note on PWM for grbl/TinyG/Smoothie right now, but being pulled in different directions, it will probably be a couple of weeks before it’s finished.
External steppers are fine - just a lot more work and unneeded for such a small machine. Unless you’re intending on a lot of metal work, most integrated controllers capable of 2.5-3A will be more than adequate for most implementations. A lot of DIYers have more money than sense 
As he’s not the most tech-savvy, you want the easiest and simplest implementation and toolchain you can offer.
A GRBL-based controller, CNCweb/Laserweb with a decent CAM app, like CAMBAM would be my choice. My 86-year-old Dad can manage that.
If you want to future-proof, get a Due or Zero and run GRBL1.1 with a 3.3v shield, then you can always upgrade to smarter firmware later on if you want. The difference between GRBL1.1 and G2Core is significant when you are pushing speed. I like the ability to send wirelessly, so the TinyG/G2 with John Lauer’s Serial Port JSON Server running on a cheap OrangePi and ChiliPeppr makes for a great and reliable path from design to production.
It’s hard to give you a ‘best’ ansewer, as there are so many choices to be made in hardware, firmware, electronics, but I think if you look at what the Shapeoko people offer, you can’t go too wrong - they are selling to first time users.
They used to use TinyG, now Grbl. They now recommend Carbide3D Motion and Carbide Create, but that’s because they are bought out by Carbide3D - it’s free and worth a look. I’ve downloaded it but not run it up yet.
For me, my favoured option is always simplicity. I like an integrated board because I can mount it on the back of the Y-axis, keeping all the control and power cables short, simplifying wiring (only DC power for the board/steppers and AC for the spindle through the power chain). My steppers aren’t huge and the DRV8825 drivers more than keep up with the power.
For software, if it’s 2.5D (sheet goods) I use SketchUp and Cambam, for 3D I use Fusion and Cambam, for Laser/raster I use Photoshop and Cambam, for vector I use Adobe Illustrator and Cambam. I have a bunch of other software like Vectric, but rarely use it, but for some it’s a better solution. (I’m a Mac/Linux guy - I only use Windows for CNC. I create on Mac, send to Win for CAM and sending to CNC).
If I was building a new machine for a new user, I would think seriously about keeping it simple. With most controllers you have the option of adding external drivers, should you need them - I never have.
I would also think about something like a Shapeoko 3 - $1500 gets you a lot of machine, with all the wrinkles sorted out and support at the end of the phone (if you are in North America - here in Asia support is something for other people 
)
I just priced up the Sphinx and by the time you add all the bits, it’s the same price as a Shapeoko. I would go with the Shapeoko if it was for a first-time user. Resale is easy for used Shapeoko should you decide to upgrade at a later date and it’s a fully-formed and supported product, rather than a DIY.
The S3 with a 33"x33" cutting area (much bigger than the Sphinx) with a DeWalt router, all electronics, controller, software in XXL size is $1800. Without the router it’s $1699
