Of course, the numbering is not canonical, but as the vertex numbers are used to display higher-dimensional constructs, they are basic for these files. You can do any orthogonal transformation on the coordinates, and the information will stay the same, as long as you retain the numbering.
In the files with the vertices also the reflectors are indicated. They are given by a vector (from the origin) that is perpendicular to the reflector. However, this vector is not normalised. So when you want to use it to do reflections be aware of this restriction. The reflectors given are not the only possible set, it is just one of them, but, whatever (valid) set you use makes not difference. Note that when you do an orthogonal transformation on the coordinates, you also need to do the same transformation on these vectors.
Some of these files use a shorthand for common constants. The most frequently occurring is tau, which is (sqrt(5) + 1)/2. Another one is sq5 for sqrt(5). I do hope that all other notations are clear.
For each section in a polytope there are files displaying the edges, faces, cells and possibly hypercells. They are all displayed using vertices only. Except in the case of the edges, there are subsections (clearly separated and marked) for the different kind of objects. Within each subsection the objects are numbered from 1 upwards. Edges are prefixed by E, faces by F, cells by C and hypercells by H. The numbering is based on the lexicographic ordering, but it is in fact pretty irrelevant, there are only a few cases where that numbering is actually used.
More information about how some specific objects are displayed is given on a separate page.
In 3D there are no data files for cells, in 4D not for hypercells and in 5D not for 5D objects. That should be clear in general. However, when compounds do occur, it is interesting to note how the different objects contribute to the compound. In that case a set of objects is preceded by a line giving some form of numbering between parenthesis. It should in general be clear how to interprete that, but in more complex cases it is explained. This will also occur when some other possible grouping is given in the information files. But play around with it. At this time not all groupings are indeed also indicated in all the data files. This is an on-going process.
If a number in a column is given as more than one number separated by asterisks, this indicates that grouping is possible amongst the objects. If the number is given between square brackets that means that grouping is possible, but very complex, the information files will give more information.
If the result column says disconnected this means in general that none of the n-dimensional objects share (n-1)-dimensional objects with others. If it says part connected there is some sharing, and in the information files this will be (should be) explored. The only difference is when the highest dimensional object is just edges, if they do not form any regular polygon, they are marked as disconnected.
Also, when the remarks column says total split this means that upto (and including) the edges level, the section splits in two or more similar subsets.
When you click on the info label you will get a closer view on the particular polytope and the related polytopes. These pages display the subsection of the table that is relevant and give more information about what is happening. In some cases all interesting information can not be displayed on that page, and in those cases there is a column notes that links to notes files that give more in-depth information about some particular sections, or some particular objects. Again, these notes files display a subtable from the base table about what is relevant to these notes.