This question has remained unanswered yet not closed, so I'll give it a go.
First of all, a clarification regarding the title. If phobias are genetic, by definition they can't be created in a generation and inherited. The term epigenetic is the one that might have led to this confusion, so I'll start by defining it.
Epigenetics is the study of changes in the expression of genes caused by certain base pairs in DNA, or RNA, being "turned off" or "turned on" again, through chemical reactions (source). Epi means outside or around, so epigenetics is not genetics as you might define it. These changes happen at a genomic level, but they don't involve a change in the nucleotide sequence. The best example of epigenetics is the fertilized egg cell or zygote, that divides and changes into all the different cell types in an organism (muscle cells, neurons, epithelium). They do this by activating some genes while inhibiting the expression of others.
The best definition is probably Arthur Riggs':
[Epigenetics is] the study of mitotically and/or meiotically heritable
changes in gene function that cannot be explained by changes in DNA
2. Can mice inherit fears applied to their predecessors?
The study you cite affirms they can. Even a third generation of mice seemed to reject the smell as well, which makes for quite shocking news:
The third generation of mice — the ‘grandchildren’ — also inherited
this reaction, as did mice conceived through in vitro fertilization
with sperm from males sensitized to acetophenone.
The problem with the study is that it failed to figure out a mechanism by which this could happen. And without knowing the molecular mechanism behind it, there is no choice but to remain skeptic.
3. A partial conclusion
This is undoubtedly a breakthrough discovery, the possibility of imprinting associations such as this in a gamete is simply astonishing. From an evolutionary point of view, it can make sense that (in some way we don't understand yet) some signals regarding experience can be transmitted to our descendants and somehow help in their survival.
The cited study belongs to epigenetics because this information cannot be stored at a nucleotide level. Unfortunately, and as I previously mentioned, without knowing the details of the mechanism we can't categorically affirm these mice inherited the information associated with a traumatic experience.
Even in a laboratory context, there are too many unknown variables we might be missing. One important one has been brought up by Martin Johnsson, and it has to do with the study using two groups of mice, one subjected to the scent and the other one not, and their descendants. If you divide individuals into two groups, you need to be sure that there are not genetic differences between the founders of the two groups. This is not clarified in the study, and could potentially explain the supposed 'inheritance' if no other reason comes to light.
So to summarize: The experiment seems to affirm it's possible. However, without a mechanism we are left with too many unknowns to consider this a proven theory.