Even though I am not sure I understand your question completely, I want to try to explain to you the genetic difference between humans and chimpanzees. I will work through your questions:
We share 98.5% genes with chimps ,so there is about one percent difference .It means we can approximately differ from them by one base pair every hundred base pairs on average.
It is correct that humans and chimpanzees differ approximately every 100 nucleotides in their DNA sequence. This is does not mean that 98.5% of the genes are shared. It means that humans have about 98.5% (more precisely about 98.8%, The Chimpanzee Sequencing and Analysis Consortium, 2005) sequence identity with chimpanzees, disregarding indels. The main part of this difference (about 85% of the 30 - 35 Mio. different nucleotides) comes from fixed differences: a fixed difference is a position in the haploid genome in which every human has, say, a G and every chimpanzee has a T. Accordingly, the site is different between the two species for every individual of the respective species. The smaller part (about 15%) comes from sites that are either variable (SNP) in humans or variable in chimpanzees, or in both (there is even a small number of very old variants, i.e. shared alleles between humans and chimpanzees, e.g. see Azevedo et al., 2015). Note that the estimates get better the more individuals of the respective species are sequenced and compared.
A human receive 3 billion base pairs from each of parents ,and has six billions in total .Do the math by using one hundred to divide six billions and then it equals sixty millions.
Due to that, The person have ,on average ,sixty-million-nucleotide differences with a chimp.
However,the professor of the online courses I attended said it's thirty-million-letter differences.
Dose that mean it's counted on haploid?Why?
Yes, it is counted on the haploid genome. I would not make any sense to count it on the diploid genome. When looking at my last paragraph you might see why: Remember, we are looking at (i) fixed differences (both copies of a diploid human genome differ from both copies of a diploid chimpanzee genome) or (ii) at variable sites. Assume you find one site in one haploid genome that fulfills either (i) or (ii). If you then look at the same site in the other copy of the genome you already know that the position is either (i) or (ii). This means that it has already been counted as a difference. If you counted it again, in the worst case, over the complete genome, you would count everything twice. In general, you would strongly overestimate the absolute difference.
Also ,I'm not sure that about one percent difference is calculated on one base pair or one nucleotide every thousand base pairs or nucleotides .Because he mentioned the both.
I do not understand this. I had a brief look at the video and heard him talking of one nucleotide per thousand. He compares the sequence difference between human-chimpanzee (98.8% overlap, approx. 1 per 100, about 1.2% fixed differences and SNPS) and human-human (99.9% overlap, approx. 1 per 1000, 0.1% SNP), pairs. Note that in the last case we cannot look at fixed differences as there are no fixed differences between human populations. The 0.1% difference are the percentage of SNP in the human genome.