The solution is simple - eliminate tenure. Then you can take a risk on a promising scientist - worst case, you fire him/her in a few years.
You may not even be able to get them at all, because there's a widespread perception that non-tenure-granting universities don't offer much academic freedom (especially in biochem, where they have strong ties with pharmaceutical companies). I could be wrong, but I don't believe these institutions have so far been successful in attracting top talent.
One hypothesis is that corporate research labs already occupy this ground well enough, so non-tenure-granting universities don't have much competitive advantage when looking for talent. If you don't care about tenure, want a higher salary, and are willing to accept the fact that management can review and direct your work, why work at a university at all? Why not Google Research or the R&D division of a pharmaceutical firm?
"...I came home in 1989...to a tenure-track job, running my own research lab at a University of London institute, where I remained until the sad demise of my career...I got tenure..."
All the same, I'm not sure how she was on the tenure track given that no UK academic appointed (or promoted!) after 20 November 1987 has tenure. Perhaps there is a more complex story with her date of appointment, or perhaps, given that Science is a US publication, she is simplifying a tenure track equivalent UK position for a US audience.
Reference for the legislation: http://www.legislation.gov.uk/ukpga/1988/40/part/IV/crosshea...
But this is a side issue (my fault, I started it!), I'd agree with your basic point that tenure is a problem. Just, it's not the only problem and it wasn't the problem in this case.