I agree that outcome is worse. However, both extremes are bad. The optimal approach would be to do a little bit more work upfront, pick a "better" thing (where "better" is at least more than the next sprint or 3!), and use that.