After devouring ORAC's link yesterday and learning a lot about TSR2 that I did not know before, a few observations...
Cost was wildly underestimated, as with TFX and JSF. Paddy Hine, in his what-if musings about what might have been if the project continued, doubts that more than 100 would have been affordable, and that high-end estimates would have precluded the F-4K program.
The overruns were being under-reported (see Wilson's comments, p16 on) and the communications between MoA/S, MoD, prime, subs and direct-to-government suppliers were rife with suppression and misinformation. ACAS(OR) was ready to scrap the project before the Oct 64 election because costs were out of control.
The airframe and engine would probably have come good, eventually - but two requirements in particular (M=2.25 vs M=1.7, and STOL/grass field) were very expensive and had been imposed without a proper cost study. As one participant remarks, there was nobody in industry to say "you can't have that". (cough TFX JSF cough)
Avionics? This would make a fascinating comparison with F-111 experience, and to consider whether TSR2 was more equivalent to the original F-111A kit, the never-worked F-111D system or the compromise Mk IIB that the C/E/F all had. It sounds a bit like -111D in which case it would be a freaking nightmare. Also, as Simmons (p107 on) notes, the best accuracy in a low-level blind delivery was several hundred yards - so the concept depended on a sizable arsenal of kiloton bombs, which nobody was signing up for in 1964-65.
With a Pave Tack equivalent and a 1980s avionics modernization, it would have been different, but until then...
Technologically, a great deal was learned. From a requirements viewpoint, one result was a more balanced set of reqs for MRCA-75/Tornado, and even Typhoon. It's interesting that while the UK/Europe have had their difficulties, it is the US where death by compounded requirements has been the rule - B-1A, B-2, F-22 and JSF.