Hey Dezie Okpoebo,
You're probably going to find that separating the historical score with absolute accurately is going to be difficult to scale. If I were trying to do this, I'd look at my new lead scoring rules and consider doing a historical one-off run on those to make sure the score is aligned with any methodology changes in score values, etc.
The demographic scoring points won't really be very time specific, so that shouldn't be too difficult to pull out - I'd just create cloned variations of the campaigns that aren't trigger based and run them on the data set. For behavioural scoring points - define a period of time that's short enough to be relevant, long enough to capture a meaningful amount of data, and run them again as non-trigger based campaigns, obviously keeping in mind that you may need to factor in the number of times something happened. E.g., if you've got "clicks email" trigger scores, you may want to consider alternatives for <5, 5-10,10-15 clicks in the past 90 days etc.
Because these would all be once off smart campaigns, I'd chuck them in a subfolder away from the ongoing trigger campaigns just to keep things tidy too. And to avoid confusion, I would consider export all current score values as a back up, and then wiping those fields.