Users with 10 reputation points hop the fence of protected questions while one with less than 15 can't even upvote an answer
"Questions should be protected when they are attracting poor answers from new users that exceed the volume which can be moderated in a timely manner."
Although the tool of protected questions is probably rarely used (I have never seen it used), its existence in the current form is implemented in a rather unthoughtful way — it either should go or be fixed. I will argue as to how the latter could be achieved.
One may already answer questions specifically fenced away to protect from poor answers including from new users, but with the same reputation one is not even considered sufficiently qualified to even judge the quality of other questions. A great divide is palpable in the level of carefulness exercised in these two situation.
If one cannot judge the reasonability, soundness or validity of information presented to them to cast a vote under 15 reputations how could they formulate a reasonable answer as that requires to first read and judge information based on which one would provide such an answer also, the same process necessary for casting a positive vote, plus the process to compose it into an answer. Clearly, casting a positive vote requires less competence as providing an answer that is not targeted by the protected question tool.
In fact, the privilege to downvote an answer is further down the road, at 125 reputations.
Since the protected question tool supposes that a lot of answers will come in for its questions targeted, there is no presumption of by heightening the fence, a question would be risked to not receive an answer. We can presume that when answers are present, one may only want to submit an answer when they believe the answers are partly or completely wrong including in missing correct information. If one is unable to judge below 125 reputations, it is unreasonable to assume that one with 10 or 11 reputations, just hopping the protected question fence, is able to judge based on the quality of other questions whether their answer is necessary when a question was protected from a potential flood of poor answers.
Since not only is one presumed to be required to be able to judge whether presumptive previous answers on a poor-answer-prone question are also poor to merit writing an answer, but also to be able to adequately compose such an answer based on a series of positive and negative judgements during their learning process necessary for giving the answer. Therefore, it seems clear to me that the fence has to be set at least >125 reputations, or the repetitional bar presumptively giving the privilege to a user to negatively judge other answers.
In fact, there are certain privileges suggesting that a lack of competence in judging questions and answers exists all the way up to 1000 reputations, the privilege to edit, inter alia answers.
How much higher may be the proper subject of debate, but I see no reason behind allowing any where lower than at least a 126 reputations.
I would argue that in the apparent rarest scenarios the tool needs to be used, it would be appropriate to allow only those to submit an answer who are presumed to be competent to submit edits without community approval.