With reference to leelee's comment, the link given is for short oligos. For a vector, this link
is more apt because linearised plasmids would have one longer flanking end that may be sufficient for digestion. However, since SmaI is not listed in the link, it may "require 6 base pairs on either side of their recognition site to cleave efficiently" as stated in the link. Therefore the long flanking end may or may not be enough for cleavage of the SmaI site.
When NEB suggests sequential digestion, it is usually to minimise star activity. In this case, it may be due to the difference in digestion temperature as well (25C for SmaI and 37C for EcoRI). Since they both have 100% activity in NEBuffer 4, I suggest you try double digestion in the buffer first (first at 25C then 37C) and see if it works [I've worked against the sequential digestion advice with no problem
]. But since EcoRI seems to have less problem cleaving a site with short flanking ends, it may be better to digest with SmaI at 25C, then add EcoRI and digest at 37C. This way, SmaI would be digesting the plasmid with two long ends flanking the restriction site and EcoRI will be left to do the "tough work" of digesting the linearised vector with the EcoRI site flanked by 3bp. Make sure the glycerol concentration is kept below 5%.