Let's get something out of the way right away. This question raised above has nothing to do with cases back from appeal. There is a different statutory provision at play when that occurs and explicit provisions have been written for that situation.
No, instead, I'm wondering if one jury can convict a defendant of an A or B felony and a different jury -- a month or two later -- decide if the defendant is a Dangerous Offender.
The answer is unquestionably no if the first jury wasn't unanimous. This is a consequence of the statutory rule that a juror who voted to acquit originally would not be allowed to find the enhancement fact. Consequently, there is no way to adequately address that acquittal with a brand new jury. You can replace a juror, yes, if you know how he or she voted, you've got a pre-approved alternate and the original juror is unable to proceed. But the same plan wouldn't work for an entire non-unanimous jury.
[Update: I think the above might need further explanation. Let's assume the jury is 10-2. The 2 jurors who voted "not guilty" get to participate in deliberations on DO, but their votes don't count. Let's assume a brand new jury is empaneled for DO. Who picks the two brand new jurors whose votes don't count? How do you decide? Rock, paper, scissors? Or by empaneling a new jury -- without any authority to do so -- does the state now get two additional jurors to vote for DO?]
So, a new jury when the old one was non-unanimous isn't feasible. But what if the jury is unanimous on guilt? Could the legislature -- though silent on the issue -- have intended that a new jury could be empaneled -- that is, an entirely new procedure -- as long as the vote was 12-0 instead of 11-1 or 10-2?
I say no. Here's why. The statute at issue -- ORS 136-770(5) -- says in relevant part:
(5) Unless the defendant waives the right to a jury trial on enhancement facts related to the defendant, the sentencing phase shall be conducted in the trial court before the jury following a finding of guilt by the jury.
Bold added, obviously. Why is the "the" important?
Because under traditional rules of statutory interpretation, the legislature is referring to the same jury. The legislature didn't say "a" jury. It said "the" jury. Think I'm being picky? Tell it to the Court of Appeals.
The legislature's use of the definite article "the" indicates that there is only one pertinent "time of judgment" at which a court may declare the conviction to be a misdemeanor. If the legislature had intended the construction that defendant proffers, that any judgment entered at any time may qualify as "the time of judgment" under ORS 166.270(3)(a), it likely would not have used the definite article "the." See State v. Nguyen, 223 Or App 286, 291, 196 P3d 40 (2008) (explaining, in the context of ORS 803.540(1)(b) that, "if the legislature had intended to allow drivers to choose among different 'fronts' of their vehicles in the positioning of registration plates, it would not have chosen the definite article 'the.'"); see also Osborn v. PSRB, 325 Or 135, 142-43, 143 n 7, 934 P2d 391 (1997) (discussing use of the definite article "the").
Apply the court's own analysis to the statute above, and I think it is self-evident that a new jury can't replace the old one if, for example, too many jurors get sick.
I would note that the same statute above does explicitly allow alternate jurors for the sentencing phase, but that language also helps the defendant. The Court has considered a procedure whereby one or maybe two jurors could be replaced, but it did not specify any procedures for replacing the entire jury. That silence is telling, given the language above, as well as the fact that there is no hint that the legislature ever imagined brand new juries would be empaneled between trial and sentencing.
In sum, for the trial court to have the authority to empanel a new jury, the legislature would have to have intended a different procedure for unanimous rather than non-unanimous juries, it would have had to have been completely silent on what the new procedure should be, and the Court of Appeals would have to throw out its prior precedent on the significance of the definite article.
So, my answer to the title question is "no."
[One final thought: what happens in a capital case? Assume some delay between the culpability and penalty phases. Have you ever heard of a brand new jury being empanelled absent the case being sent back from the Oregon Supreme Court? In the history of capital cases in Oregon, has that ever happened? I don't know, but I'm guessing it hasn't. Why not, do you think?]
[One more final thought. I might be making this too complicated. When does jeopardy attach in a jury trial? When the jury is sworn. So, in a case where one panel decides guilt and another decides culpability, then jeopardy attaches . . . twice? Isn't there a name for that? Jeopardy times two? Jeopardy Plus? Jeopardy Deux? Well, anyway, something for the habeas attorney, if it gets that far.]