You shouldn't need to do it in two steps, I have used tailed primers of over 100 bp in length successfully.
Assuming that you are wanting to then clone the tagged gene your primers will be something like : 3-6bp RESITE 3bp ATG HATAG 3bp FLAGTAG 3-6bp GENEOFINTEREST (with or without the ATG on the gene of interest, approximately 20 bp of the gene). You can also add a consensus Kozak or Shine-Delgarno sequence upstream of the start codon to enhance expression of the gene in eukaryotic and bacterial cells respectively.
Yes, you don't need to reverse anything for BLAST2seq, it will find the sequence even in reverse orientation. ClustalW AFAIK not.
(but any sequence viewing software like Chromas or FinchTV should be able to reverse the whole plot including sequence, this is generaly not only good for searching within sequence, but also if you want to print sequencing plot from reverse primer in forward orientation)
In our lab we use Sequencher for aligning our sequences because our university has a site license for it, but you can align the sequences without purchasing software by putting your non-coding strand in a reverse-complement calculator (like http://www.bioinform...s/rev_comp.html). Then you can use the output in any sequence alignment site you want (BLAST or ClustalW or whatever, although they may handle converting to reverse complement themselves; I didn't check), or you can just look at your forward sequence, find a series of 8 residues or so and search for them in the second sequence. Once you know how they match up, you just have to copy and paste the extra sequence from your reverse primer reaction onto the end of your forward sequence.
Does that make sense? I'm sure there are good sequence alignment sites out there that would do this for you, but I'm not familiar with them.