Why not just enlarge to various magnifications using, say, 8x10 papers and see the results for yourself? It couldn't require more than a halve dozen prints and one session. Then you could compute the size the total print would be.