The first one is a smaller board with fewer parts, but it requires SMD soldering and programming a CPLD. Looks like the glitchwrks one is all throughhole. It depends on your capabilities and experience which one you might want to tackle.

If it was me I'd use a throughhole CPLD for easiest board assembly.

it's of course your prerogative to make one if you wish, And I will cheer you on. But I am curious as to your motivation though as assembled ones are not that expensive next to the risk of making your own and the additional effort.

I have a blue lava one which works well. no ide port though. and the risk of backwards insertion is always having me worried.

Sphere,

Half of the fun is ordering the PCB, waiting 4 weeks, ordering the parts, burning them while soldering, fitting some the wrong way etc! 😁 Otherwise it's boring! 😁

SMD appeals me - gives me some opportunity to do SMD soldering. I have no problem with that. It's just that the project seems a bit "dead in the water"?

I see, makes sense. I love challenges, too. It's just the opposite that appeals me. 😆
I find it fascinating to make as much as possible by my self.
That's why I love through-hole, discreet circuits, etc.

But SMD per se isn't bad, I think. There are reasons for an old school fan to give it a go, too.
Balloon missions (high-altitude ballooning), single-sided PCB designs, R/C models etc..
Also, SMD parts don't have to be small per se. Big SMD components also existed.

In this case, I'm a bit surprised that both designs use a flash/eeprom still.
Why does the CPLD version not something like a microcontroller/SRAM/serial EPROM etc?
That chip surely could assist interfacing modern RAM to the bus..

Edit: The link to the benchmark is dead ("buffers=99").
Here's an alternative link to it.
https://www.lo-tech.co.uk/wiki/DOS_Disk_Tester