Given how polling in IA has been the past several cycles, Trump is probably further ahead here than this.
The guy is likely right, this poll is using a sample where dems and republicans are tied (35% each) while republicans have generally a small advantage in Iowa. Among the people who voted in 2016 you have 42% Clinton voters and 48% Trump voters while Trump won Iowa 51/42, thus this poll is likely underestimating Trump a bit.
The "who did you vote for" sample is rarely, if ever, identical to the last election.
