From the math side of things it doesn't look like a sound idea.
Let's say you are doing the exact thing you are talking about. Right off the bat of the shoe you put 4 units out. Lets assume your unit is $10 bucks just for example. Let's say in the most optimal conditions in a 2D (S17, DAS, etc) the house edge is .2%.
Every time you are placing your 4 unit bet ($40) you have an expected loss of 40 x .002 = .08 so for every time you place this bet you will lose 8 cents. Not really that big of a deal. However, on the average the count will fluctuate right around 0. Let's say that in an hour of play you play 100 hands of your 4 unit bet at a negative expectation. You are now giving up $8 an hour and to someone with a unit of $10 dollars thats a significant portion. (100 hands at around 0 is a rough estimate)
What happens if the running count is around +1 or -1 or just at 0? Do you keep your same bet? I don't have my frequency table with me, but I think you will find you are staying at your 4 unit bet very often and will most likely have a negative expectation from your system.
Of course if it goes negative you drop it, and if it goes positive you raise, but I don't think you should go immediately to 8 units if it goes positive. If the true count increases to +2 or +3 a maximum bet (8 units) will have your bankroll swinging so much that you will see some very high days and very low days, and from experience the low days really eat you up especially if you do not have discipline.
I think its important to play your best game and not give up any advantages such as the one you are considering. Too many people believe that everyone is on to them and that they have to use all these elaborate betting schemes or 'donkey' plays that they are giving up real money that hurts their bottom line.
Play the game you know and play it to its fullest, do not give the house any advantage, for they give you none. If you want to build up comps then work on rat holing chips over time. Your losses will look bigger and your wins smaller. Also if you are able, maintain short sessions or pay attention to when the pit changes shifts and work around that time.
EDIT: Seeing that a few people replied I also agree with the letting it ride mentality. This is a fantastic cover play because gamblers will push their luck all the time. Another good move is if you end a show at your 8 unit bet, maybe throw out a 3-5 unit bet on the first hand of the next shoe, if the count is negative or zero drop to one unit. I find personally that just saying things like "I'm going to press my luck here." or talk about how the last shoe was "Lucky". But most of all I do not think you should put yourself at such a risk with your 4 unit bet all the time from the start.
I hope I have answered your question.