The T45 does have its advantages, mainly that its radar is very high giving it a better height of eye against incoming sea-skimming missiles. It's not really debatable, it's just science.

 

That said, when you compare the radars which version are you comparing. The SPY-1A is not the SPY-1B/D is not the SPY-1D(V). The D(V) is a very interesting radar. Yes SPY-1 is a PESA vice Sampson's AESA, but I'd put money on the SPY-1D(V) if for no other reason than it has build on 4 decades of experience. 

 

Now which is the better ship? IMO, DDG-51 beyond a shadow of a doubt. The T45 is reliant upon the Aster missile which hasn't been tested against supersonic threats, AFAIK. The DDG-51 uses the tried and true Standard Missile family. The DDG-51 carries more missiles, can use the ESSM and has Mk15. The T45 has Aster 30 and 15. It can carry the Mk15, but doesn't. DDG-51 can defend against ballistic missile threats; the T45 cannot.

 

If I'm going into battle, I'm taking the DDG-51-class. Then again, I am a bit biased towards it.

Radar to radar in my opinion AN/SPY1 is better because its got 4 seperate arrays and full contious coverage. Sampson is higher than AN/SPY1 and therefore has an increased horizon but the lower amount of arrays is a cost cutting measure as the machinary to spin the the whole thing surely cannot be much less than the weight of another 1 (or 2) array needed for full coverage. However comparing the actual platform I'm not so sure the DDG-51's are not equipped with a secondary air search radar (as far as a I was aware) unlike the T 45 which has the Alenia Marconi Systems/Signaal S 1850M long-range 3D radar how much this impacts performance I dont know...

One thing I would point out with regards to SAMPSON is that the publically released information states that the rotation is something like 30 revs per minute - such that no piece of sky is left for more than one second on average - albeit this can be shorter or longer depending on the electronic sweep of the AESA radar beam.

 

So I would contend that in that respect, the disadvantage of rotating two arrays is minimised if not totally negated.

 

Of course what we are discussing is merely the front end of a large integrated system.