Some splits are offensive (increase your winning expectation) and some are defensive (decrease your losing expectation). You parlay into a defensive split and you have done yourself no favor. Doubles are sensitive to the count. Making them willy nilly after parlays is a recipe for disaster. Ina high count you will have great opportunity doubling many hands you might not and getting better results on your doubles. In a poor count basic strategy doubles may be risking double to decrease your expectation. Another recipe for a parlay disaster. You will get longer streaks in a poor count, both winning and losing streaks. High counts tend to have short streaks because the dealer gets good cards as well. The difference is you get a bonus on the more frequent BJs and can choose to vary your plays due to the high count. The dealer must play the same regardless of the count.

Besides the math of your idea being totally against you, the nature of the runs in the game will have you betting your highest bets most often in poor situations. This will further hurt your outcome than the math shows. The math says if you bets are made independent of advantage you expect to lose your total of all your bets times the house edge. Obviously the best strategy is not to play and you expect to lose nothing. The next best option is to flat bet the minimum. Of course if there was a way to know how the advantage shifts as the cards are dealt...