My reaction is that this will be a nice beast if you never need a real 68000.
I do, since I don't want the A500 to be the "portdown target" of already slow software. I foresee drivers and file systems never being made 68000 compatible because of this.
It seems a tiny bit strange without the 68000; the one advantage of owning a real A500, that all games and demos run perfectly, is replaced with a solution that could be a little more compatible than WinUAE, perhaps a lot less.
It could be a WHDLoad machine for those who have expanded chipmem. For the others, an A1200 with fastmem (or ACA1230) is still a better solution to run all OCS games.
Was planning to buy two, but I can't really use this to 'enable' my OCS demomachine until 68000 compatibility is guaranteed. Lucky I kept my SCSI interface!
Lots of text and seems a bit harsh, but this is the only criticism I have for it. It will certainly be a cheap enabler for all those A500s in the closet. As such it's inevitable that we will get more Amiga users!
But I still think an "accelerator off, my old A500 back so I can run this stubborn program please" would have been a good idea.
He decided to lock 1 IDE port to an onboard CF slot, so presumably it's meant to be 'the harddisk' and adding files and backing up is done via network. At least the network port is outwardly accessible, so if there's built-in network drivers and software to transfer files hopefully the chicken-and-egg problem is neatly solved.
If that happens, that alone is a selling point for the ACA520 - in a few minutes your unused A500 is back in business without any hassle!
Edit: someone on a1k asked about transfer speeds; I saw no reply by Jens but it's reasonable to assume that it will be "Buddha IDE speeds for a stock A1200 with fastmem with 0 wait state". If someone has such a rig
and post the results in the Performance Gather 2011
thread, we'd have a good estimate