The main point to consider is probably that of success. The relevant questions here are:
- Can you address the justified criticism?
- Can you convince the editor or even the reviewers that the unjustified criticism is really unjustified?
- Do you think that you can convince the editor to consult the opinion of yet another referee?
- How much of the rejection is based on predominantly opinionated aspects, such as relevance and scope (e.g., “not enough physics”)? Apart from some exceptional cases (e.g., if you can actually add something to the manuscript that increases the relevance), such criticism is difficult to address.
- Are the reviews criticising the same or different aspects?
Some examples:
- You have three reports. All criticise a lack of relevance. They may or may not criticise other aspects. – Resubmitting is very likely a waste of time. Lack of relevance is a killer argument and most of the time, swaying a referee means convincing them that they erred in their assessment, which even in academia few people are capable of admitting. And you probably have to sway at least two referees. Moreover you have to convince the editor that you properly addressed the issues, so there will even be another round of review. It’s unlikely that the editor will consult additional referees.
- You have two reports. The first one only criticises the lack of relevance but has no other major point of criticism. The second one finds the manuscript relevant but strongly criticises a technical aspect, which you can address. – The second referee can probably be easily satisfied. Unless the first referee has overlooked something essential, it will be difficult to sway them. However, the referees are one-to-one on this matter and by addressing it, you may convince the editor. Another possibility is that the editor consults the opinion of a third referee.
- You have two reports. Both are negative but outrageously sloppy: For example, they criticise claims you never made, base their criticisms on misunderstandings they could not possibly have noticed as such or ask you to discuss alternative approaches which you did already discuss. – You can easily address the criticism and possibly sway a referee that is more awake during the second round. If you make your case clear, you may also convince the editor to consult new referees.
Another aspect to consider is that of time:
- If at least some referees will be used again in the second round of review, this round is usually going to be faster, as it reduces or eliminates the time spent on finding appropriate referees and waiting for referees that decline to review. Moreover a referee that is already familiar with a manuscript needs less time to review it.
- If you manage to convince the editor that some of the referees from the first round were not properly doing their job, this may lead to a speed-up of the next round.
- If a major hurdle is the editor accepting your rebuttal, you waste little time if you fail, as such a decision usually happens quite quickly.
- Depending on the constraints of alternative journals, submitting your manuscript there may require some rewriting work.
Therefore, even if you consider a resubmission to be less likely to succeed than a submission to the next journal, it may result in the least average time to publication.
Finally, you have to consider the number of journals that exist on the desired level and that are appropriate for your manuscript in terms of scope, length and possible other constraints. If there are only few such journals, you are more likely to exhaust your options and thus resubmitting is more likely to be worth the effort and time.