Chip Nataro and I have been having a back-and-forth conversation on this topic for about 18 hours now and neither of us likes where we are, so I am opening up the thread to the larger community.  However, I will preface this with, I think the question is poorly written and the answer they want is simple, when in fact, the answer is complex.

Here goes. Spessard and Miessler, 2nd ed, chapter 5, end of question problem 5.5:

The Fe-C distance in ferrocinium+ is 6 pm longer than in Fc, but the Co-C distance in Cobaltocinium+ is 6 pm shorter than that in Colbatocene.

Ok, I think we can all agree and understand the difference between the 19e CoCp2 with an electron in the eg* (Co-C antibonding) orbital.

But Fc+, 17e- complex, I would agrue that the dz2 orbital is non-bonding and removal of an electron is unlikely to account for much bond-length change. However, the text claims that dz2 is *slightly* antibonding (though in the MO diagram, they show it as *slightly* bonding, which would actually make sense; removal of an electron from a bonding orbital would increase the Fe-C distance...)

Chip and I were able to find a few references (a theoretical paper from 1975 suggests that the HOMO in ferrocene is a triply occupied e2g orbital which you might then expect to J-T distort, DOI:10.1007/BF00551403).

Of course, I asked this question without thinking about the answer and now I need to write a key...

if you have any insights, thoughts, comments, please add them below!

Adam