Yes sorry, you get 50% from each parent. The “flip a coin 23 times” I was referring to was grandparents. You get 23 chromosomes from each parent, one half for each pair, to make up a full set. the 23 you get from your father, say, come partly from grandpa an partly from grandma, since he came from the same process. (If dad gives you a Y congrats, you’re a boy! You got paternal grandpa’s Y. If you’re a girl, congrats too, you have your paternal grandma’s X. ) The odds that you would get exclusively one grandparent’s 23 is astronomical.
And as pointed out, this ignores the possibility of Chromosomal crossover - Wikipedia - which again, is not likely zero. Plus the risk of transcription errors, where there are mistakes in copying the DNA during cell splitting.
As for the question about “what percentage is my contribution?” this is hard to say based on limited information. Are the regional indicators based on a few specific chromosomes or spread over several? are both of a pair Broitish? You can only give one of a pair to your offspring. (with Meosis crossover, you could have a chromosome that’s part English and part European but as a result of a crossover generations ago… or it could be your fault. ) But you cannot contribute more percentage than you actually have - there’s no doubling up.
Some base it off the http://www.internationalgenome.org/ project, with added proprietary data, typically using principal components analysis (PCA).
PCA is a useful tool but it is also sensitive to population counts, and these random markers that are chosen by the above product, or the chip manufactures are not typically exclusive to a particular area, but they are just common there.
While they try hard to correct for this if there was a marker that was in lets say Finns and Tongans, but they had more Finns as samples you would cluster more there and these changes happen over a continuum and populations mix more than most people realize.
These tools are far more useful at a population or a clade level, and less useful at an individual level just because the differences between one human and another are actually very small. As an individual you have a high chance of being more similar to someone on another continent then an individual in the same region.
If you say trace your lineage back to Charlemagne 40 generations ago, he is but 1 of your 1,099,511,627,776 direct ancestors at that level, and this is why if you have European ancestry you are a direct descendant of anyone alive in 900AD who has a living descendant. Obviously there weren’t 1,099,511,627,776 people alive during Charlemagne’s time and our family trees are more like family wreaths.
There are some ways to group populations and it is useful, and I myself really enjoy looking into this aspect myself but the problem is that there is no real segmentation that can be called English on a biological basis on an individual to individual basis just some markers that may group that way that can describe clustering of sub populations that happen to be culturally English.
Remember that English is cultural and culture is not biologically inherited.
These services that try to provide regional linage are best thought as entertainment and on the larger group level approximations. They are useful, but they select very specific locations from a very tiny part of the DNA as a whole and then simply decide what they know from a very very tiny sampling of the population. This is still useful for tracking migrations and inferring history that is even more vague in archaeological evidence but not nearly fine grained enough or reliable enough to produce hard answers for an individual.