Interesting question.
I've just got out my old TDK booklet on their audio and video cassette range around 1990.
TDK list the E180 as having a total thickness of 19.5 u meters. But the C90 audiocassette is listed as having a total thickness of only 12.5 u meters!
So as far as I can tell, the VHS E180 videotape was actually thicker based than the C90 audio cassette.
Interesting too that their shorter videotapes like E 120 and E 60 used exactly the same tape stock, not thicker ones. Just shorter lengths.
You'd think that if there was a durability advantage from thicker tape bases they'd have gone for it in the shorter lengths where there was room in the shell for it. But to be fair, maybe with the delicate setup of a videotape where tolerances are pretty tight, a different thickness tape might have caused more problems than it was worth. I'm only guessing.
I think the main stresses on the videotape were related to the rotating drum and things like staying in pause mode for extended periods when you could literally scrape the coating off the base. Possibly the reason for the apparently tougher tape than with audiocassettes.
In any case those extra stresses on the videotape would not apply in a deck with no head drum.
So that's my reasoning. If a tape can survive a rotating head drum, it can surely survive stationary heads. I dont think we should assume that just because it was a videotape as such that it was fragile. Rather, video required the rotating heads and that was probably the achilles heel of the thing.
I guess the same would apply to DAT tapes and these days to mini DV camera tapes which I use myself, but, I admit, with some trepidation! Three years and so far, so good, but I transfer the data as soon as I can just in case.
Interestingly, DAT tapes are notoriously problematic and arent trusted as archival media. Again, a rotating head configuration.
You mention 7.5ips or 15ips as possibly stressful for the tape. I doubt it would be. For example good quality standard audiocassettes can happily be high speed duplicated at 30ips.
And that's duplicating in the actual cassette shell. Commercial pre recorded recordings could be high speed duplicated serially on pancakes running at over 100 times play speed! So the limiting factor for speed was not the tape but the mechanical abilities of the cassette housing or shell. Cassette tapes couldnt be duplicated much faster than x100 because above that the tape would to "fly", meaning the trapped air between the tape and the record head would lift the tape off the head just enough to reduce recording clarity! We're talking a tape speed approaching 200ips!
From everyday experience we know that VHS, Beta, Umatic and other tapes survive REW and FF at speeds far in excess of the design play speed. And that's not capstan controlled, the preferred method for ideal tape wind.
I dont have a VHS spec list but found an old Beta one. On an L500 tape the average rewind speed (which in a Beta machine meant fully engaged with the head drum) was 46.8cm/sec which translates to well over 15ips (38cm/sec)
Anyway, my thoughts FWIW.
Cheers Tim
Akai did this in the mid 80’s with their MG Series12-tracks (14-tracks really) using a proprietary tape shell similar to Betamax. It was expensive at the time and you could only record for 10 minutes at high speed (7.5 ips). I was surprised back then that the idea didn’t catch on with other manufacturers and improved upon, but I think less expensive systems like the Tascam 388 and then 8-track on cassette from TOA and Tascam helped doom the Akai approach. 
