I designed it to work that way because I assumed what you describe was developed for the traditional financial markets which opened and closed every day, and therefore didn't always have a continuous seam of trading between each candle. Perhaps my take on it is traditionally wrong.
I suppose the gaps are typically a lot larger when a market closes overnight, but that method is used on all charts in the financial world, regardless of timeframe. In a fast moving or illiquid market, there may be gaps even on 1-minute candles. By treating the first trade in that time period as the open of the candle it gives an accurate picture of the true range that traded for that candle.
Hey, just a note that I've updated the logic to start recording open prices the way you suggested. It doesn't make sense to count the price of the last trade in the previous interval as the open price, when I'm not counting it towards the volume.
You can't really tell the difference unless you're viewing 1/3/5/15 min charts, but I'm glad to have it drawing candles in the more conventional way.