Diff signal names before comparing two waveform dumpers' cost

When benchmarking two dumpers on the same design, a scope-based dump call (e.g. $fsdbDumpvars(1, top.u_core)) also dumps the module's input ports, clk and rst included. A hand-listed or tap-based capture of the module's internal leaves doesn't. The clock is the busiest signal in the design, so a cost ratio between the two is biased even though the scopes look the same. Before reporting a ratio, export both sides' variable names (VCD $var lines vs the other tool's signal list) and diff them with comm. Then either equalise the sets or label the ratio with the extra signals.