(Problems with a mailserver earlier this week, outgoing mail was lost,
so I'm re-sending. My apologies if this is a duplicate.)
My suggestion (worth every cent you paid for it) is to use a whole damn
computer on the other side of the HD controller.
Construct an electrical interface of the simplest possible hardware to
get the job done (electrical interface, buss termination as required,
etc) and do the whole job in C or even Perl.
Simulating the block structure of the disk would be trivial in software.
The hardware handshake and DMA junk would be where the work is --
regardless of the level simulation abstraction, but with say two EPP
parallel port cards you could have the handshake nicely defined in (more
readable than yelling schematics) code.
A mediocre gigaHz cpu has gotta keep up with all but the fastest disk
controller of yore.
And then later when EIDE and serial ATA are as obsolete as core, your
simulator is more likely to port to whatever the crapyp hardware of 2020
is.
(The assumption is that translating C, perl, whatever will be easier
than soldering.)