Hi Damilola,
the RIS Export utility that you find on the present site is a separate tool that we have created starting from the facilities we had in EPPI-Reviewer 3. This allowed us to take advantage of all the import filters that we had written over the past years. In EPPI-Reviewer 4, the import mechanism is brand new and as such it requires us to write new filters for every file format that people might need to use. Currently, there are only 4 filters in EPPI-Reviewer four, mostly because people are happy to use the "Ris Export" utility and very rarely ask us to write new additional filters.
In practice, this means that there is no requirement to use the RIS export utility: if your import file is in one of the four formats supported natively, then there is no need to use the additional tool. In any case, it is important to check that the import did recognise the references correctly: we can't make any guarantee, as the number of different formats out there is enormous, and even within the same format there are frequently small differences or "dialects". This is why we wrote a new system for EPPI-Reviewer 4: it allows us to write more robust filters, able to recognise most different dialects correctly. The RIS export utility is more sensitive to small differences in the format, and this is why you will need to make many choices to select the right filter.
PubMed, on the other hand, can be searched within the main application; also, any reference-list file written in PubMed format can be imported directly. Both methods are equivalent from the functional point of view. I am not sure about what you mean with "added the correct right and end tags and it worked just fine": did you have to modify text files created by PubMed in order to import them correctly?
As for importing the full text: I'm afraid this has to be done manually for each reference. PubMed does not normally hold the full texts, instead, it frequently provides links to the full-text versions on the publisher websites. Wether a single user will indeed have access to a particular full text depends on his or hers affiliations/subscriptions and is inherently unpredictable beforehand. For this (and many other) reason(s), it is not possible for us to design a mechanism that will automatically fetch the full text versions in any reliable way.
Since obtaining the full text version of an article may also have a cost, systematic reviews usually include a first round of screening based on Title and Abstract only. In this way the number of articles for which the full text needs to be obtained is reduced to a (hopefully) maneagable amount.
I hope this helps,
Sergio