This is really something. If I'm understanding correctly, your computer is in complete control of the transport?
I see a lot of potential in this, and I'm wondering...how hard would it be, assuming you had the assistance of a DAW and SMPTe (or some sort of timecode), to record different parts of a song on different parts of the tape?
Here's what I'm thinking:
You have a TSR-8 and a BR-20T both slaved to your computer (hypothetically), plus CLASP, a DAW, and an analog mixer. Also, you have found some way to bypass the half-track deck's erase head. (possible?) Now, you want to record more than 8 tracks, but you'd rather not do an internal bounce, so you record 7 parts of the song (and 1 for timecode), and then using your DAW to comunicate to your tapedeck, you record another 7 parts for the song further along the tape so as not to erase the first parts. Then, using the computer-controlled transport and the timecode on the tapes, you mix down the first 7 tracks to the BR-20T (in real time through your analog console...keep the signal chain analog), then go back and mix down the 2nd set of tracks over the first set, being perfectly in sync and remembering that the erase head is disabled. Get my drift? Is this possible?
Or, for instance, say you had a multitrack machine (that had could do 30, 15 and 7.5ips) slaved to your computer. You want to record the drums and bass at 7.5ips, electric guitar and vocals at 15ips, and acoustic guitar and a string quartet at 30ips, all in the same song and mixed down to 2-track at 15ips. Beneficial at all?