The reference to EP1 by EPKM is a problem for enablement.Withdrawing EP1 solves this, but EP1 is the first application for the signal.
I did the exam in the office under exam conditions, and I admit I missed the complication - I read the Guidelines, and simply suggested withdrawing EP1. To get the broadest invention patented (signal), I suggested claiming priority of EP1 with EP3.
I thought of it afterwards, so I asked a couple of knowledgeable people and they didn't know for certain either.
The Case Law book A.I.2(b) says:
In T 737/90 it was explicitly stated that a reference to another document can only be taken into account if the document referred to can be unambiguously identified and the relevant addressees have ready access to it. This depends solely on the facts of the case.
The public must be able to find it on the publication date of EPKM - that is clear from the guidelines.
It is clear that EP1 is due to be published 1 month before EPKM - therefore, if you do nothing, EPKM will be enabled.
If you withdraw EP1 prior to publication, and do not claim priority, then the public cannot find it so that EPKM is not enabled.
If EP3 claims priority of EP1, you are in a grey area. Sure, EP3 is published then 1m before EPKM, but EPKM refers to EP1. If you type in the application number of EP1 in the "application" field in espacenet, you will get no hit. But if you type the EP1 number in the "priority" field, you will find EP3.
In the OJEPO case law supplement of 2007, you find:
T341/04: the question was whether a referenced document, which could be unambiguously identified at the date of filing of the document containing the
reference by its document number, but which document itself was "missing" in the sense that it was not available at said date of filing, could be "taken into account" for the purpose of Article 83 EPC by relying on information present in a family member of the referenced document.
In this particular case, the board answered in the affirmative. The EPO would have easily retrieved the document, as would a member of the public, availing himself, if necessary, of the professional skills of a librarian on or after the (international) publication date of the application that matured into the patent in suit.
I have read the case and it was very similar - a published family member claimed priority from the referenced filing.
As I said, I didn't know the sufficiency test in such detail. I suspect not many candidates knew it either. I don't like them using minute details of patent law for the exam - people can get very frustrated because they cannot figure it out in the exam. It could also be that they did not anticipate this complication when they made the exam.
However, the DII marking is usually constructed so that if you miss a couple of things, but do well on the rest, you can still get more than 30 points. Most people probably saw the issue, and got some points for discussing enablement and withdrawing EP1. If a lot of people struggle with it, they can even reduce the weighting in the marking.