The mere presence of a conditional isn't a determining factor, but the context of the conditional can be a contributing factor in determining whether the threat is true or not. See a more recent publication from 2003 for a better understanding of those factors. I definitely wouldn't rely on anything from the 1600s as a reliable source.
Back in the old days a threat was essentially the same as an assault, but nowadays we often view threats and assaults more distinctly and have crafted more specific laws to address them. Just as an "arrest" was once used to describe a seizure of any kind regardless of the duration, more modern case laws have established distinctions between arrests (full-custody seizures) and detentions (temporary, short-duration seizures).
Since most jurisdictions in the US now have a separate statute for addressing threats in order to distinguish them from assaults, we have to read the exact verbiage of each state to determine whether it covers the particulars of the threat. Here are a few examples. Note the similarities and differences in each:
Arizona: Threatening or Intimidating
Texas: Terroristic Threat
New York: Menacing
As you yourself pointed out, simply making a threat isn't necessarily mala en se, and there are times where the law explicitly allows for people to issue threats, particularly when it comes to defending themselves, another person, or their property. For example, in Arizona, ARS 13-407 allows a person to issue a threat of deadly force to defend their property against trespassers, but does not allow them to actually use deadly force to do so. This means that Arizonans who find themselves in a situation where they feel it necessary to issue a conditional threat to an intruder ("leave my property right now or I'll shoot you"), must understand that the threat is empty unless the intruder shifts his focus from mere trespassing to harming a person. (The bad guy doesn't need to know this though, and all the better if he doesn't.) Even though the threat is empty according to the issuer's knowledge, it is still a threat, albeit one that the issuer is legally entitled to make. In this situation, the threat to use deadly force is decriminalized, not because it contained a conditional, but rather because it was explicitly authorized by law.
In practice, many jurisdictions, especially ones with high crime rates, have arbitrarily decided to triage threats cases based on whether they include a conditional statement or not. After all, if you have 1,000 cases a month and a team of only 30 prosecutors, you have to prioritize the cases that are easiest to prove. With that in mind, many prosecutors offices have provided direction to their law enforcement partners to only submit cases that include unconditional threats. Those agencies in turn train their officers to ignore any threats cases involving conditional threats, perpetuating the misunderstanding that a conditional automatically negates a threat as being a true threat.
There are literally thousands of cases where offenders have had their convictions upheld for crimes involving conditional threats. Just think of robbery/carjacking: A man walks up to a victim in a parking lot and says "give me your car or I'll kill you. Despite containing a conditional threat, there isn't a prosecutor or jury in the country that wouldn't instantly recognize this as a true threat and an element of robbery/carjacking.
Conditional Intent to Kill is Enough for Federal Carjacking Conviction.
