Two things -- not major points, but to ignore them is inappropriate.
1) When you do estimates of parlay usage for projects like this, you should take into account the disadvantage of using parlays locked into one sportsbook at one point in time. Those disadvantages are not negligible. You are paying a price for using non-optimal numbers, as the vast majority of the time, you are not getting the best of available numbers vis-a-vis shopping at eight or 10 books over time.
2) The gripe I have with the use of the term "EV" applies here. "EV" is supposed to be a precise term grounded in relatively precise estimations of probability, not somebody's ballpark best guess based on their personal experience and how much coffee they drank that day. In other words, either preface the "EV" with the word "subjective" or "my ballpark estimate of the" or something. Please don't plug in the term "EV" as a ballparking term without adopting the pronoun "MY estimate of the 'EV'" or something similar. "EV" should not be used when there's a good dose of subjectivity involved, which is usually the case. The problem from a language perspective is that mathematicians would interpret "EV" as a math term. Gamblers looking for an excuse to gamble use it as a synonym for "myy opinion."
I don't disagree with your use of the bonus strategy for the working off free play, but anyone relying on math to make these decisions should recognize the not-so-hidden cost of doing business via parlays of any kind. You are very rarely getting optimal numbers, and those non-optimal numbers have a cost. Now if you're betting $50 or $100 parlays, no big deal, but if you're dealing with significant amounts of money, it's important to recognize the non-optimal numbers as such.